712 



Fishery Bulletin 102(4) 



and orange roughy (Hoplostethus atlanticus; Beamish, 

 1979; Archibald et al., 1983; Mace et al., 1990). These 

 historical examples of fishery collapses necessitate that 

 age validation be achieved before age and growth infor- 

 mation is applied to management. 



In the last decade, radiometric age validation has 

 been applied successfully to over 20 species of rock- 

 fishes and other marine teleosts (Burton et al., 1999; 

 Kastelle et al., 2000; Andrews et al., 20021. The most 

 common technique uses the disequilibria between two 

 radioisotopes, radium-226 ( 226 Ral and lead-210 ( 210 Pb), 

 present in the otolith (Bennett et al., 1982; Smith et al., 

 1991). Radium-226 is a naturally occurring radioisotope 

 and calcium analogue that is incorporated from the 

 surrounding seawater into the aragonitic crystalline 

 matrix of fish otoliths. Radium-226 decays through a 

 series of short-lived radioisotopes to 210 Pb. Because 

 the half-lives of these isotopes are known, the ratio of 

 activity between them ( 210 Pb: 226 Ra) gives a measure of 

 elapsed time since the initial incorporation of 226 Ra into 

 the otolith (Campana et al., 1990). Radium-226 decays 

 very slowly (a 1600 year half-life) in relation to 21 "Pb 

 (a 22 year half-life), allowing the activity ratio of these 

 radioisotopes to build into secular equilibrium (1:1 ra- 

 tio; Smith et al., 1991). Based on this relationship (also 

 referred to as ingrowth), the 210 Pb: 226 Ra activity ratio 

 is suitable for age determination in fishes up to 5 half- 

 lives of 210 Pb, or approximately 120 years of age (An- 

 drews et al., 1999b; Campana. 2001). This approach is 

 therefore ideally suited to the blackgill rockfish, whose 

 longevity has been estimated at almost 90 years (Butler 

 et al., 1999). 



The objectives of this study were 1) to estimate age 

 from otolith growth zone counts, 2) to describe growth, 

 and 3) to validate the annual periodicity of growth 

 zones used to estimate longevity for the blackgill rock- 

 fish with the radiometric aging technique. An ancillary 

 objective was to create a reliable predictive relationship 

 between average otolith weight and estimated age for 

 use as a timesaving tool in the management of this 

 species. Growth zones quantified in sectioned otoliths 

 were used to estimate age, and growth was described 

 by using the von Bertalanffy growth function. Final age 

 estimates were directly compared to radiometric ages to 

 evaluate agreement between the two methods and ulti- 

 mately were used to validate age estimation procedures, 

 age-at-maturity, and longevity for this species. 



Materials and methods 



Approximately 1210 blackgill rockfish sagittal otoliths 

 were available for this study. Otoliths were collected 

 by National Marine Fisheries Service (NMFS) person- 

 nel from commercial vessels in 1985 at ports along the 

 California coastline (Long Beach to Fort Bragg), and 

 during NMFS research surveys from 1998 to 2000 from 

 central California to the Oregon-Washington border. 

 Thirty-two juvenile blackgill rockfish, collected from 

 spot prawn traps along the central California coast, were 



provided by Robert Lea of the California Department of 

 Fish and Game (CDFG). Fish total length iTL; cm or 

 mm), catch area (port or geographic location), and otolith 

 weights (right and left. 1985 samples only) were pro- 

 vided. Otoliths were first considered for age estimation 

 i sectioning!, and the remainder were reserved for radio- 

 metric analysis. Otolith weights (left and right, male 

 and female) were measured to the nearest milligram 

 and compared with £-tests to determine if significant 

 differences in mass existed between sides or sexes. 



Estimation of age and growth 



Based on previous aging studies and the need to con- 

 serve samples for radiometric analysis, approximately 

 310 otoliths (25% of the collection) were assumed to be 

 sufficient for age estimation. The left otolith from 5 to 

 30 fish, depending upon the number available in each 

 50-mm size class (ranging from 100 mm to 600 mm), 

 was randomly chosen by using a basic resampling tool. 

 Otoliths were thin-sectioned and mounted onto glass 

 slides. Approximately 50 otoliths were damaged in the 

 sectioning process, leaving 260 otoliths available for 

 age estimation. 



Sections were viewed by three readers under magnifi- 

 cation (25 and 40x) with transmitted or reflected light. 

 Each reader obtained age estimates by inspecting all 

 available growth axes, choosing the most discernible 

 axis, and reading it three times consecutively. A growth 

 zone (here termed an "annulus") was defined as one pair 

 of translucent (winter-forming) and opaque (summer- 

 forming) bands. A final age, based on each reader's most 

 confident estimate, was chosen. Precision between and 

 within readers was compared by using average percent 

 error (APE; Beamish and Fournier, 1981 1, index of pre- 

 cision (D) and coefficient of variation (CV; Chang, 1982). 

 Percent agreement among readers was also calculated. 

 Reader 1 (author) determined the final age estimate for 

 each section as described in Mahoney (2002). Ages that 

 could not be confidently resolved (through re-examina- 

 tion or discussion) were removed from analysis. 



Length and age estimates for males, females, and sex- 

 es combined were fitted to the von Bertalanffy growth 

 function (VBGF). A small portion of juvenile samples 

 (/j =16) were included in each function. Because there 

 was strong agreement between facility aging techniques 

 (MLML and NMFS, La Jolla. Butler et al. 1999), ad- 

 ditional aged samples were added to strengthen the 

 VBGF and age prediction models (ra=119). Estimates of 

 age at first, 50%, and 100^ maturity were calculated 

 by inserting existing size at maturity data (Echeverria. 

 HIST ) into the VBGF and solving for age (t). 



Age prediction, age group determination, 

 and core extraction 



Campana et al. 1 1990 ) was the first to circumvent the 

 assumption of constant 226 Ra uptake throughout the 

 life of the fish by eliminating younger growth layers 

 from adult otoliths, leaving just the oldest layers of 



