Adam et al.: Dynamics of Thunnus obesus and T albacares in Hawaii's pelagic fisheries 



223 



Discussion 



The size- and site-specific attrition model described in this 

 study is new and potentially applicable to other fish species 

 where release and recapture data meet the model require- 

 ments. What is required are the size and geographic posi- 

 tion of releases and recaptures. One difficulty encountered 

 related to the quantity of release data that was available 

 for analysis. Because releases were stratified over 1-cm size 

 class cohorts to reliably track their growth over time, larger 

 numbers of releases would be required to have reasonable 

 numbers in the cohorts. Thus we assumed that all tags 

 were released at some arbitrary time zero. 



Attempts to estimate size-specific transfer rates were 

 unsuccessful because of poor convergence of the numeri- 



cal estimation procedure. Size-specific transfer rates were 

 poorly defined in the data sets because of the low number 

 of recaptures in the relevant strata. It is, however, trivial 

 to incorporate size-specific transfer rates in the model and 

 use the same procedure to estimate the transfer rates by 

 the size classes under consideration. 



The growth of yellowfin and bigeye tuna in our model is as- 

 sumed to follow the von Bertalanffy growth function for the 

 entire lifetime of the cohort. However, Lehodey and Leroy,^ 



'' Lehodey, P., and B. Leroy. 1998. Age and growth of yellowfin 

 tuna {fhunniiK alahcarcst from the western and central Pacific 

 Ocean as indicated by daily growth increments and tagging 

 data. Working paper 12. Eleventh Standing Committee on 

 Tunas and Billfish, Secretariat of Pacific Community, HP D.5, 

 98848 Noumea Cedex, New Caledonia. 



