Le GUEN and SAKAGAWA: APPARENT GROWTH OF YELLOWFIN TUNA 



180 r 



160 - 



140 



PRESENT STUDY 

 (APPARENT KNOWN AGE) 



q: 

 O 



120 



100 - 



3 4 



APPARENT AGE (YEARS) 



Figure 9. — Comparison of growth of yellowfin tuna from 

 the Pacific and Atlantic Oceans. Curves were adjusted 

 to a common base of age 1.5 years = 60 cm long and 

 were estimated, except for that of Le Guen et al. (1969), 

 from data reported in various studies. 



method — requiring length measurements at 

 equal time intervals — Moore was able to use only 

 16 out of his 25 observations. We recalcu- 

 lated Loo and K, using the Fabens' procedure 

 and the 25 observations reported by Moore (his 

 Table H) . The estimates, L« = 191.9 and K = 

 0.036, differ slightly from those of Moore and 

 are very similar to our estimates for Atlantic 

 yellowfin tuna. 



Le Guen et al. (1969) estimated growth of yel- 

 lowfin tuna from Dakar, Pointe-Noire, and both 

 regions combined, based on modal progression 

 of length-frequency samples (Table 7). Their 

 samples were identical to some used in our study, 

 but their estimate of growth for combined re- 

 gions is slightly lower than ours ; the difference 

 in estimated lengths for ages 2 through 5 years 

 is 2.8 to 4.3% less (Table 4). Part of the dif- 

 ference is in the method of analysis. The esti- 

 mates by Le Guen et al. were based on mode se- 

 lection from predorsal length distributions, and 

 the lengths of size groups were not assumed to 



be normally distributed. Predorsal lengths were 

 then converted to fork length; whereas in our 

 best estimate predorsal length was converted to 

 fork length by a log function before frequency 

 distributions were analyzed, and the lengths of 

 size groups were assumed to be normally distrib- 

 uted. Furthermore, Le Guen et al. assumed that 

 the date of birth of fish of each year class of a 

 region was the same and accordingly ages were 

 assigned to size classes; such an assumption was 

 not made for our estimate of K and I^ ; but for 

 obtaining to we assumed that yellowfin tuna of 

 60 cm long are 18 months old. Nevertheless, the 

 difference is insignificant in view of the fact that 

 there is considerable variability in observed 

 mean lengths at age (Figure 7). 



DISCUSSION 



It is obvious from the results that estimates 

 of growth of yellowfin tuna are quite variable 

 and largely dependent on the method of anal- 

 ysis. Both the length-frequency and scale meth- 

 ods are based on various assumptions that are 

 not always satisfied. For example, the assump- 

 tion in the length-frequency method that size 

 groups represent age groups, and the age groups 

 are formed once a year, i.e., hatching within a 

 short period, or season, is not completely sat- 

 isfied for yellowfin tuna, since spawning occurs 

 over several months (Matsumoto, 1966; Le Guen 

 et al., 1969; Richards, 1969). Nevertheless, in 

 many areas, as in the eastern Atlantic, there is 

 generally a peak month of spawning (Le Guen 

 et al., 1969) that can create a size group discern- 

 ible in size-frequency distributions in later dates. 



The scale method assumes that the scale marks 

 are formed at regular intervals. So far, this as- 

 sumption has not been satisfactorily verified for 

 yellowfin tuna, although Yabuta et al. (1960) and 

 Yangetal. (1969) have indicated that the marks 

 formed every 6 months. Furthermore, because 

 yellowfin tuna generally spawn over an extended 

 season, the age at first annulus formation is not 

 the same for all individuals of a year class. The 

 back-calculated length at age I may therefore 

 be questionable. It is surprising, however, that 

 the observed lengths at age are remarkably sim- 

 ilar for studies based on the scale and length- 



185 



