Labelle et al Determination of age and growth of South Pacific albacore 



659 



120 



4 5 6 7 



Time interval (years) 



Figure 4 



Growth of south Pacific albacore from an initial length of 40 cm as pre 

 dieted by models fitted to length-frequency and vertebral-ring-count data. 



— . 1. 



: 500 days 



length-frequency modes are semestral, rather than an- 

 nual features. We then compared each set of predic- 

 tions to the observed recapture lengths. 



The growth models derived from length-frequency 

 and vertebral-ring-count data, assuming that the modes 

 and rings are annual features, appear to be consis- 

 tent with the tag return length increment data 

 (Fig. 6). Deviations of observed 

 from predicted recapture 

 lengths are both positive and 

 negative and are generally 

 within the range of expected 

 individual variation. In con- 

 trast, the growth model based 

 on the assumption of semes- 

 tral length-frequency modes 

 overestimates the recapture 

 lengths. On this basis, the tag 

 return length increment data 

 lend support to our initial 

 assumptions concerning the 

 length-frequency modes and 

 vertebral-rings. 



Discussion 



Estimates of the VB growth parameters for 

 south Pacific albacore based on the analy- 

 sis of length-frequencies and vertebral-ring- 

 counts were generally within the ranges of 

 values reported for albacore from other re- 

 gions (Table 5). The only exception was the 

 L. estimate obtained with MULTIFAN, 

 which was lower than all other estimates 

 reported. This could be partly due to the 

 fact that the data set used for the 

 MULTIFAN analysis did not include length- 

 frequency samples from longline catches, so 

 the largest size classes were not well repre- 

 sented. Nevertheless, we do not attach any 

 biological significance to this or to the dif- 

 ferences in VB parameters estimated from the two 

 data sets in this study. Despite differences in the pa- 

 rameters, the predicted growth patterns based on the 

 two analyses are remarkably similar. This often noted 

 feature of the VB growth model (e.g.. Knight, 1968; 

 Francis, 1988) highlights the potential pitfalls of com- 

 paring growth only on the basis of VB growth param- 



1 



5 



© 

 C 



1.1 



1 



.91 

 .8 

 .7 

 .6 

 .5 

 .4 

 .3 

 .2 

 .1 



>500 days 



55 



60 65 70 75 

 Mid-size (cm) 



85 90 55 60 



65 70 75 

 Mid-size (cm) 



80 85 90 



Figure 5 



Individual growth rates of tagged south Pacific albacore for two time-at-liberty categories. 

 Mid-size is the fork length halfway between the release and recapture lengths. 



