TTTt r^ \ \ \ \ ^ \ 'I \^^ 



' >' -SO" 160° 170° 180° 170° 160° 150° 140° 130° ^< 



,,_-^HAWAIIAN 



-^^^ FISHERY- 



w/////Mv/////MMmw//M. mmmmm^ 



JAPANESE LONGLINE FISHERY 



POFI 



EXPLORATORY 



180° 



_J 



170° 



I 



150° 



_1_ 



Figure 1. --Fishing areas from which data were obtained. 



randomly from the population. " Since most of 

 our samples consist of a number of small catches 

 from many fishing stations well separated in 

 space, they probably fulfill this requirement. 



Even large, well-distributed samples may 

 not be representative of the size composition of 

 a migratory population. Although fishing effort 

 in the vicinity of the Hawaiian Islands is roughly 

 constant throughout the year, there is a marked 

 increase in the catch of yellowfin during May to 

 September, indicating a probable migration into 

 and out of the fishery (Otsu 1954). If yellowfin 

 available to the fishery at different times have 

 different growth rates, the analysis will not 

 describe growth accurately. 



We must also recognize that no matter 

 how we 11 distributed in time and space the samples 

 may be, longlines do not catch small fish and 

 thus provide little or no information on the early 

 life. Moreover, in the frequency distributions 

 of larger adults the modes are closer spaced and 

 more difficult to distinguish than in the smaller 

 juveniles. 



Another complication is the protracted 

 spawning period of the yellowfin. Although the 

 most active spawning period may be during the 

 sumnner months, equatorial y e 1 1 o w f i n with 

 running milt have been observed the year around 

 (Shimada 1951a). It has been suggested by 

 Schaefer and Marr (1948) that "several batches" 

 of eggs are spawned over an extended period of 

 time. June (1953) notes that "individual fish 

 /yellowfin/ mature more than one group of ova 



and have several spawnings during the same 

 season in Hawaiian waters. " His 1950 obser- 

 vations showed that spawning extended from 

 around the middle of May to the end of October. 

 Wade (1950) indicated that the spawning of 

 yellowfin tuna extends from May to August in 

 the Philippine region. As the result of such a 

 prolonged spawning period, nnodal groups may 

 be extended and ill-defined, msdcing size- 

 frequency distributions difficult to analyze. 



In addition to these difficulties, male 

 yellowfin taken by the longline are larger than 

 females. Many of the nneasurements used here- 

 in are not accompanied by sex data, and al- 

 though failure to separate the sexes may not 

 cause serious errors in interpretations, it 

 will certainly increase the variance of modal 

 groups and make it harder to define them. It 

 is also possible for modes to shift with tinne, 

 not through growth, but simply by changes in 

 the sex ratio in the sampling area. As will be 

 shown, this danger does not seenn to apply in 

 the present instance, although caution must be 

 exercised in interpreting changes in the modal 

 groups of unsexed samples. 



METHODS 



The lengths of nearly all equatorial 

 samples were taken in nnillimeters (some to 

 the nearest whole centimeter) nneasured from 

 the tip of the snout with jaws closed to the 

 median portion of the caudal fork (with the 

 fleshy flap depressed), as described by Marr 

 and Schaefer (1949). These length measurement? 



