1, The following has been revealed in 

 compariion of the length frequency by fishing 

 area and season. 



A. As far as the samples caught in the 

 same season are concerned, modes 

 of the length frequency are, as a rule, 

 found in an agreement with each other, 

 almost regardless of fishing areas 

 (Fig. 1). 



B. When the modes formed in different 

 seasons are compared, variance in 

 their presentation is remarkable de- 

 pending on fishing areas. However, 

 it has been observed that character- 

 istic features prevailing throughout 

 the period under study indicate domi- 

 nant modes, some formed by smaller 

 fish caught in the western areas, and 

 the other by larger fish in the east. 



This can b e interpreted in other 

 words that the small sized fish were 

 as abundant in the western part just 

 as the large fish in the east (Fig. 2). 

 Judging from the facts stated above, 

 it is hardly possible to consider that 

 the population of the big-eyed tuna mi- 

 grating to the fishing grounds under 

 discussion would consist of more than 

 two groups belonging to entirely 

 independent races. 



C. In regard to annual discrepancy in the 

 modes, attention has been drawn to 

 the fact the modes tend to fornn with 

 frequency occuring every other year 

 (Fig. 3). 



2. As probable causes of the annual 

 fluctuation in the modes three hypotheses have 

 previously been established by Nakamura, 

 Kamimura and Yabuta 1): 



(i) Difference in annual propagation; 



(ii) Difference in annual growth rate; 



(iii) Difference in annual course of 

 migration. 



In the examination of hypothesis (i), a 

 series of assumption (A, B, C) have been in- 

 troduced in connection with successive trans- 

 formation of the modes which is likely involved 

 in the course of growth. The assumption has 

 been derived from length frequency by season 

 (Fig. 4) as tabulated below: 



r^ishing Season 

 Assumption 1948 1949 1950 1951 1952 



-D3 



-B, 



As the result we have reached the conclusion 

 that hypothesis (i) is more reliable than the 

 others so far as progression of the nnodes in 

 the way assumed above appears reasonable. 

 Whereas, hypothesis (ii) can not be, at least, a 

 decisive factor influencing the mode progression. 

 Because it is difficult to draw rational interpre- 

 tation of the phenomenon from this hypothesis. 

 On the other hand, there is some room where 

 hypothesis (iii) may be accepted as reasonable 

 one, provided that the fish should take the same 

 migratory route every two years. 



3. On the basis of the above assumption, 

 as it seems to be most feasible for the present 

 study, ecological mechanism influencing the 

 biennial frequency has been considered. In 

 consequence, notable fluctuation taken place in 

 abundance of age groups every second year is 

 found to be likely a major factor supporting this 

 particular phenomenon. At the same time, it 

 has been pointed out that there is more or less 

 disagreement between the modes formed in 

 different seasons when examined in the light of 

 age-group, which might likely be ascribable to 

 the ecological mechanism as well. 



4. For the purpose of proceeding the study 

 further on the basis of the assumption (A, B, C), 

 consideration has been paid on two factors: one 

 being the stage from which the biennial frequen- 

 cy might have developed; the other, ecological 

 mechanism involved therein. It is almost 

 certain that the initial stage for the frequency 

 would have occurred before the fish developed 

 to age groups ai_3. However, an alternate 

 question that remains to be solved lies in whether 

 the biennial frequency was resulted from; 



a) causes accumulated in the course 

 growth; 



of 



b) o r a cause which had a decisive effect 

 on the fish at a certain stage of the 

 development. 



Judging from general circumstances for 

 marine life, the latter cause (b) is more likely 

 to take place than the former. In addition, there 



21 



