Pearson and Shaw: Age determination errors for Anoplopoma fimbria 



139 



Table 11 



Percent agreement between number of hyaline zones counted by three age readers and the number which should have been pres- 

 ent. Also shown is the effect of removing the count of a hyaline zone which formed in the summer and which should not have been 

 counted as an annual mark. 



Reader 1 



Reader 2 



Reader 3 



Original 



Corrected 



Original 



Corrected 



Original 



Corrected 



24% 



44% 



36% 



44% 



20% 



Effect of ages on stock assessments 



Crone et al. (1997) noted that one of the problems with 

 stock assessments of sablefish is that the size at 50% sexual 

 maturity is between 55 and 67 cm ( age 5-7 ) and that there 

 is considerable variability in the these estimates. Further, 

 they noted that there has been difficulty in determining 

 age-specific selectivity because of problems with the ages 

 used in previous assessments. Crone et al. (1997) further 

 noted that there is a considerable discrepancy in ages 

 among the age determination laboratories on the west 

 coast. Finally, the model used to perform stock assessments 

 has estimated that in order to obtain a good fit with the 

 data, the actual level of aging error should be higher than 

 has been reported. The lack of reliable age data has been 

 used to criticize stock assessments. 



Age and length at sexual maturity has been found to 

 vary substantially by depth (Fujiwara and Hankin, 1988a). 

 Fujiwara and Hankin found that both males and females 

 had a length of 550 mm for the length at 50% sexual matu- 

 rity in shallow water (<600 meters ). In depths greater than 

 600 m, the size at 50% sexual maturity was 450 mm for 

 males and 500 mm for females. To determine age, they used 

 sectioned otoliths and methods that may not have been 

 directly comparable to the methods used in other studies 

 or the methods used in the present study; nonetheless, they 

 found that both males and females matured at a younger 

 age in deeper water. Saunders et al. (1997) also reported 

 differences in length at maturity related to depth and loca- 

 tion of capture. Methot 1 found that ontogenetic movement 

 into deeper water for spawning was more closely related to 

 age than size. If sexual maturity is more closely related to 

 age than length as suggested by Methot, then unreliable 

 ages may explain the variable maturity schedule for sable- 

 fish. In our study, fish were captured over a 900 nmi range 

 at depths from 200 to more than 1000 m. If depth is related 

 to growth of sablefish, then it is possible that the different 

 morphometric types of otoliths observed in our study may 

 also be a function of depth. If depth is responsible for the 

 morphological types, it also suggests that reliability of ages 

 may be a function of the depth at which the sablefish are 

 found. Further, if depth influences growth, a fish which 



1 Methot. R. D. 1995. Geographic patterns in growth and 

 maturity of female sablefish off the U.S. west coast. Unpubl. 

 manuscript, 39 p. NOAA, NMFS, Northwest Fisheries Science 

 Center, Seattle, WA. 



changes its depth over time, may exhibit different patterns 

 of growth throughout its life which would further compli- 

 cate the problem of determining reliable ages. 



Potential sources of error in this study 



This study used sablefish caught in the southern part of 

 the sablefish range. Many species show latitudinal varia- 

 tion in growth (June and Reintjes, 1959; White and Chit- 

 tenden, 1977; Leggett and Carscadden, 1978; Shepherd and 

 Grimes, 1983; Pearson and Hightower, 1991). It is possible 

 that the results of this study do not apply to the northern 

 portion of their range. 



Another potential source of error in our study is the effect 

 of tagging on the growth of the sablefish. MacFarlane and 

 Beamish ( 1990 ) found that tagged sablefish grew slower 

 than untagged fish. If this is true, then the results of this 

 study are much more difficult to interpret. MacFarlane and 

 Beamish did not use OTC and as a result they based their 

 ages on conventional aging methods. If they had injected 

 the fish, it would have been interesting to note whether 

 the ages for the fish in their study would have been inter- 

 preted differently. If fish do grow differently after tagging, 

 many age, growth, and validation studies will need to be 

 re-evaluated. 



Conclusion 



Obtaining accurate ages, with reasonable precision, for 

 sablefish is very difficult. Previous aging studies of sable- 

 fish have obtained results similar to ours, even when the 

 readers knew how many annual marks should have been 

 present (Beamish et al. 1983; MacFarlane and Beamish, 

 1995). We found that some fish lay down two marks a year 

 and others may not lay down any. We also found that certain 

 morphological types of otoliths may be indicative of slow 

 growing fish and others may be indicative of rapidly grow- 

 ing fish (assuming otolith growth relates to fish growth). 



The fact that agreement among readers or with the cor- 

 rect age consistently ranges between 30% and 45% sug- 

 gests that this imprecision may be inherent in sablefish 

 aging. A substantial fraction of the population may not be 

 able to be reliably aged: some otoliths do not appear to 

 grow and others grow very rapidly, laying down prominent 

 summer hyaline zones that even experienced age readers 

 cannot differentiate from winter hyaline zones. 



