NOTE Piner et al : Use of oxygen stable isotopes and the 1983 El Nino to assess accuracy of aging Sebastes melanops 



555 



Table 2 



Results of the retrospective analysis that estimated the 

 statistical significance of the magnitude of the average 

 negative residual from the years 1983 and 1985. Assay 

 results from otolith material formed after the cutoff year 

 were removed from the randomization analysis. 



Results 



The period 1980-90 was characterized by an isolated 

 and historically strong 1983 El Nino event (Fig. 1), that 

 resulted in a 1-2°C increase in water temperatures 

 along the Oregon coast. Average summer water tempera- 

 tures declined slightly over this period. The o ls O values 

 measured in each fish resulting from this period (Table 

 1) showed strong patterns that indicated temperature 

 differences both within fish (between years) and between 

 fish (same year). In addition, many fish contained trends 

 in 6 18 across time (Fig. 2). Precision of the reported 

 6 18 measurements ranged from 0.01 to 0.079cc (SD). 



The residual patterns (Fig. 3A) showed that anoma- 

 lously warm conditions existed in otolith material cor- 

 responding to those of 1983 («=9, P=0.0338) and 1985 

 (ti=7, P=0.0409). Both old (ages 11-12) and young (ages 

 6-8) fish appeared to have similar temporal patterns 

 of residuals; however in older fish this pattern shifted 

 by 1-2 years toward more recent years (Fig. 3B). The 

 year-specific averaged residuals of the age 6-8 fish de- 

 picted a single anomalously warm year corresponding 

 to 1983. The anomalously warm year in the age 11-12 

 fish was 1984-85, thus explaining the significance re- 

 sult in 1985. The results of the randomization test were 

 not sensitive to the exclusion of data from the four most 

 recent years (Table 2). 



Discussion 



The location of the anomalously warm signal in 1983, 

 in the youngest and likely the more accurately aged 

 fish, supports the hypothesis that the 1983 El Nino can 

 be detected by using oxygen stable isotopes. From this 

 analysis, we concluded that the break-and-burn aging 

 method is accurate on average but has a potential ten- 

 dency toward underaging fish >10 years. 



Confirmation of the annual banding pattern in the 

 ololiths of other Sebastes species has been accomplished 

 by using a variety of methods. Woodbury (1999) con- 

 firmed the accuracy of age assignment in widow {Se- 



bastes entomelas) and yellowtail (S. flavidus) rockfish, 

 using the change in growth increment width associated 

 with El Nino. Piner et al. (in press) has used bomb ra- 

 diocarbon to confirm the annual pattern of otolith band- 

 ing in canary rockfish (S. pinniger) and has reported a 

 possible underaging bias for older fish. Andrews et al. 

 (1999) used radiometric age determination to confirm 

 the longevity of long-lived species. However, a larger 

 study on black rockfish with stable isotopes is necessary 

 to conclusively determine age estimates accurately and 

 potential underaging bias. 



The 1983 El Nino was chosen for the present study 

 because it was one of the strongest recorded in the 

 century (Sharp and McClain, 1993). Warm water condi- 

 tions associated with the 1983 El Nino were sufficient 

 to slow growth (MacLellan and Saunders, 1995; Wood- 

 bury, 1999) and alter reproductive patterns (VenTresca 

 et al., 1995) in species occupying similar geographic 

 ranges. In contrast, this study attempted to indirectly 

 measure the environment experienced by black rock- 

 fish without the need to infer changes to biological 

 processes. Nevertheless, our results appear to support 

 the conclusions of MacLellan and Saunders (1995) and 

 Woodbury (1999) that the anomalous oceanic conditions 

 in 1983 are identifiable. 



The analysis of model residuals rather than raw iso- 

 tope ratios is more appropriate because of the obvious 

 d^O temporal trend in some samples. Otolith process- 

 ing difficulties also may have contributed to the trend. 

 The opaque region of the otolith decreases in size with 

 increasing age. The narrowing of the otolith region as- 

 sociated with older ages made precise sampling more dif- 

 ficult and may have resulted in accidental sampling from 

 otolith material outside the opaque region. The sampling 

 of otolith material from outside the opaque region may 

 have contributed otolith material formed in cooler waters 

 in contrast to the sampling of areas of the otolith associ- 

 ated with younger ages. The increasing trend in & H was 

 not explained solely by the decreasing temporal trend of 

 summer water temperatures. However, an additional com- 

 ponent of that trend may be the result of age-dependent 

 fish movement to cooler waters that are deeper or more 



