274 



Fishery Bulletin 91(2). 1993 



(variability in LJ the James' method MSE was the smallest; for 

 Simulation 3 (i.e., both observational error and variability in 

 LJ the unweighted Fabens' method MSE was the smallest. 



It is clear from these simulation results that unweighted or 

 weighted Fabens' estimates of von Bertalanffy parameters can 

 be significantly biased. It is also clear that standard errors for 

 James' estimators tend to be larger than standard errors for the 

 unweighted and weighted Fabens' estimators. 



Estimates from sablefish tag-recapture data 



Sablefish is a species characterized by rapid growth at young 

 ages followed by extremely slow growth at older ages. In tag- 

 recapture data, it is common for sablefish to show negative 

 growth increments (Figs. 2,3,5,6). The exact reason for this ap- 

 parent shrinkage is uncertain. However, Sasaki (1985) noted 

 that shrinkage was usually reported for larger male (>60cm) 

 and female (>70cm) sablefish. This suggests that such occur- 

 rences are due to the combined influence of slow growth and 

 measurement error. Furthermore, Beamish & Chilton (1982) 

 noted that the freezing of commercially-caught fish could cause 



actual shrinkage. Interestingly, Beamish & 

 Chilton (1982) also noted that data from 

 freshly-caught fish measured by trained bi- 

 ologists occasionally showed shrinkage, indi- 

 cating that shrinkage can occur in the ocean 

 environment. Considering these factors, all 

 recoveries, including those having negative 

 growth increments, were used in the sable- 

 fish data analyses. 



Although this species is extremely difficult 

 to age (Kimura & Lyons 1991), age-determi- 

 nation criteria have been validated using 

 oxytetracycline tags "(Beamish & Chilton 

 1982) and radioisotope methods (Kastelle 

 1991). Because of the difficulty of directly de- 

 termining ages for sablefish, estimating 

 growth using tag-recapture data is of par- 

 ticular interest. 



10 30 50 70 90 



LENGTH AT RELEASE (CM] 



o 



O 



o 



o 



o 

 o o 



i , . .- ''■' ' •■' 



o 

 c9 



r 



o 



SO 



1 2 4 6 8 10 12 14 

 TIME TO RECAPTURE (YR) 



Figure 2 



Length-frequency at time of release (top) and re- 

 lationship between time at liberty and growth 

 (bottom) for male sablefish in Gulf of Alaska. 



