Table 6. — Average jiumhcr of larval tunas under 10 m.- of ocean surface and number of schools sighted per 10-mile run for 



the legs of the offshore surveys 



within an allotted timo, investigation of schools was 

 discontinued if the response to chumming was un- 

 favorable. Consequently, the specific identity of 

 many of the schools was not determined, and our 

 examination of the relation between larval and adult 

 abundance was in terms of the aggregate of all tuna 

 species. 



In comparing the number of schools sighted per 

 10-mile run and the average number of larval tunas 

 under 10 m.^ of ocean surface for the legs of the off- 

 shore surveys and for the inner, middle, and outer 

 7>5-mile sections of these legs, no significant correla- 

 tions (Spearman rank correlation) were found (tables 

 C) and 7). For the entire offshore survey area, the 

 averages for both adult tuna schools and larval tunas 

 were highest during either Hi\IS-43 or CHG-38, 

 slightly lower during HMS-45, and lowest during 

 CHG-35. A similar pattern was found in the aver- 

 age of zooplankton volumes. The variations of all 

 three averages are illustrated in figure 5. 



Table 7. — Average number of larval tunas under 10 m."- of 

 ocean surface and number of schools sighted per 10-mile run 

 for 75-mile sections of the legs of the offshore surveys 



INFERENCES CONCERNING 

 SKIPJACK SPAWNING 



Inferences about spawning based on the size of the 

 larva upon hatching have been discussed by Matsu- 

 moto (1958). He hypothesized that skipjack arc 

 2.5 mm. or less at hatching, that the eggs and larvae 

 are planktonic and therefore subject to dispersion by 

 (■urrcnts, but that their displacement from the 



spawning site would be relatively insignificant unless 

 the currents were exceptionally strong. Larval 

 skipjack have been taken throughout the area 

 around the Marquesas Islands (fig. 4). Most of the 

 catch consisted of larvae between 3 and 4.5 mm. 

 long, so we may assume that they had hatched re- 

 cently. Since the currents around the Mar(|uesas 

 Islands are suspected to lie weak (Sverdriip, John- 

 son, and Fleming, 1942: p. 702), these larvae could 

 not have drifted very far from the spawning sites. 

 Thus, skipjack spawning appears to occur tiuoiigh- 

 out the sampled area. 



Matsumoto (1958) has reported larval skipjack 

 catches from long. 180° to 120° W., and on the basis 

 of records of larvae and juveniles taken in the Pliilip- 

 pine Islands (Wade, 1951) and off the coast of Cen- 

 tral America (Schaefer and Marr, 1948; Mead, 1951) 

 and of juveniles caught in the Marshall Islands 

 (Marr, 1948), he has indicated the possibility that 

 skipjack spawn throughout the equatorial waters of 

 the Pacific. Subsequently, Klawe (1903) noted the 

 occurrence of larval skipjack in the eastern tropical 

 Pacific. Matsumoto (unpublished) recently ob- 

 tained larval skipjack from areas west of 180°, par- 

 ticularly around the Marshall Islands and the east- 

 ern part of the Caroline Islands. Capture of larval 

 skipjack in localities still farther west in the Mari- 

 anas and Palau Islands was reported by Yabe, 

 Yabuta, and Uoyanagi (1903). These records con- 

 firm Matsumoto's hypothesis of the transoceanic 

 distribution of larval skipjack in the Pacific. 



Matsumoto (1958) also has reported the north- 

 south distribution of larval skipjack as extending 

 from lat. 25° N., to 1432° S. in the central Pacific. 

 The southern limit now may be extended to at least 

 lat. 18° S. 



Table 8 shows the months during which larval 

 skipjack have been taken in northeastern French 

 Oceania on various cruises by vessels of the Bureau 

 of Commercial Fisheries. They were captured in 

 all months except July, in which no sampling was 



6 



U.S. FISH AND WILDLIFE SERVICE 



