Francis et al.: Age and growth estimates for Polyphon oxygeneios 



239 



1 - 



in 



a 8 



a. 



0) 



E 2 



3 





 10 



8 



6 



4 



2 







Poor Knights 

 n=16 



Cook Strait 

 n=57 



Oamaru 

 n = 65 



40 



60 70 80 



Total length (cm) 



100 



Figure 11 



Length-frequency distributions (at tagging) of tagged and recaptured 

 hapuku used in ttie GROTAG analysis of growth rates, n = sample 



experiment included only five hapuku, only one of 

 which was at liberty for more than one year Further- 

 more, all five fish would have been immature or just 

 maturing at the time of injection. Further validation is 

 required, especially for older, mature hapuku. 



An independently derived von Bertalanffy growth 

 curve (based on MULTIFAN analysis of length-fre- 

 quency data) agreed well with the otolith-based 

 length-at-age curves over the age range 0-12 years 

 but indicated faster growth rates for hapuku aged 

 12-14.5 years (Fig. 10; Table 5). Our previous expe- 

 rience has shown that MULTIFAN gi'owth curves are 

 generally reliable for younger age classes, which of- 

 ten exhibit discernible length modes, but are not re- 

 liable for older age classes, which usually lack any 

 modal length structure. This is because MULTIFAN 

 underestimates the number of older age classes 

 present in the data and consequently overestimates 

 the length-at-age of the older age classes (Francis 

 and Francis, 1992; Francis, 1997 ). Therefore, the dis- 

 crepancy between the MULTIFAN and length-at-age 

 curves for the older age classes in Fig. 10 probably 



arises from bias in the former This interpretation is 

 supported by the near-linearity of the MULTIFAN 

 growth curve and the implausible MULTIFAN 

 growth parameters, especially L^ (Table 2). 



