FISHERY BULLETIN: VOL. 69, NO. 3 



Sixty-six percent of the stomachs came from 

 billfishes captured near Hawaii' between July 

 1962 and April 1966. These stomachs were used 

 previously in an early life history study of al- 

 bacore, Thunnus alalunga ( Yoshida, 1968) . Most 

 of these billfishes were caught within 37 km (20 

 nautical miles) of the main islands. A few were 

 caught as far as 740 km (400 nautical miles) 

 from Oahu. Thirty-four percent of the stomachs 

 came from billfishes caught in the South Pacific 

 between lat 5° and 32° S, and between long 135° 

 W and 179° E (Figure 1). These were caught 

 on longline gear by boats operating out of Amer- 

 ican Samoa between January 1964 and July 1966. 

 The fishery has been described by Otsu and 

 Sumida (1968). 



Arrangements were made with the crews of 

 several boats to have billfish stomachs collected 

 on the fishing voyages. Each cooperating crew 

 was provided with a stainless steel tank, form- 

 aldehyde solution, labels, and collecting bags. 

 The crew was paid 50 cents for each stomach. 



In the laboratory at Honolulu, all tuna and 

 tunalike specimens were sorted from the stomach 

 contents and identified. Skipjack tuna were 

 identified by skeletal characters (Godsil and 

 Byers, 1944). Standard length was recorded 

 for all intact specimens. Following a technique 

 used earlier (Yoshida, 1968), a method was de- 

 vised to estimate the standard length of frag- 

 mentary specimens. Relations were determined 

 between the standard length and the length of: 

 (1) the complete vertebral column (41 verte- 

 brae), (2) the precaudal vertebrae (vertebrae 

 1-20), (3) the caudal vertebrae (vertebrae 21- 

 41), (4) Ist-lOth vertebrae, and (5) 21st-30th 

 vertebrae, based on 77 intact juvenile skipjack 

 tuna specimens from around Hawaii and the 

 South Pacific. All the relations appeared to be 

 linear and straight lines were fitted to the data 

 by the method of least squares. Combining the 

 samples from Hawaii and the South Pacific 

 should not adversely affect the results. A plot 

 of the data for all five relations did not indicate 

 any differences between the North Pacific and 

 South Pacific samples. A covariance analysis 

 applied to the relation between the standard 

 length and the length of the complete vertebral 

 column for Hawaiian and South Pacific juvenile 





^^ 



y^ 



W///////////7//, 



'■'///?/^'/7//, V/???///^ 



♦t^i-':*-'^"--'^* h'fi'^'-'-:y^^^ ,, MARQUESAS 



^ •.?t SAMOA IS. L "' WMmm 



Figure 1. — The location of capture of billfishes (shaded 

 area) and the distribution of juvenile skipjack tuna 

 (dots) in the South Pacific. 



skipjack tuna confirmed the lack of significant 

 diffei'ences between the samples. No significant 

 differences were found in the regression coeffi- 

 cients {F = 0.105; df = 1, 73) and in the in- 

 tercepts {F = 0.053; df = 1, 74). The re- 

 gressions of the standard length on the various 

 vertebral segments are presented in Table 1. 

 The lengths of most of the specimens were 

 estimated by using a suitable regression. For 

 22 ""f of the specimens the relative position of 

 the fragments could not be determined, and the 

 regressions were not used. For these specimens 

 the standard length was estimated by comparing 

 the average length of the vertebrae in the frag- 

 ment with the average length of the vertebrae 



546 



