102 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



To investigate differences between localities, 

 we may compare the food of yellowfin of 1,000 to 

 1,299 mm. fork length (approximately 43 to 92 

 lb.), captured in the Line Islands, with the food 

 of a group of the same size range from the Phoenix- 

 Islands. In figure 6, we used average volume per 

 stomach for each of the major food groups as an 



u 



-^ 



I 

 z 



e 



<n 



<r 

 u 

 a. 



i 



55 

 S3 

 45 

 40- 



ss 



30 

 25 

 20 



20 



15 



20 



15 



10 



N = I 21 N= IS 



TROLL 



CAUGHT 



N* 22 N = 44 



POLE -AND- LINE 



CAUGHT 



N*I5 



LINE ISLANDS PHOENIX ISLANDS 



Figure 6. — Differences in availability of the major food 

 categories between the Line and Phoenix Islands, as 

 indicated by average volume per stomach, for yellowfin 

 1,000 to 1,299 mm. fork length. The troll- and pole-and- 

 line-caught fish were captured within 10 miles of land; 

 the longline fish within 25 miles of land. N = number 

 of stomachs. 



index to the availability of those organisms in 

 each locality. As shown in figure 6, the most 

 apparent locality differences lay in the propor- 

 tionately greater average volume of crab larvae 

 (megalops) found in the Line Island yellowfin, at 

 least in respect to the surface-caught fish, and 

 the greater average volume of fish in the Phoenix 

 Island yellowfin. During the 2-year period cov- 

 ered by this study, we failed to encounter in the 

 Phoenix Islands any swarms of crab larvae such 

 as occurred in the Line Islands during May and 

 June, 1950 and 1951. The data indicate that 

 locality differences may be of considerable im- 

 portance and should be considered in an evalua- 

 tion of the food of the yellowfin. 



Not only are differences in the availability of 

 food organisms between localities evident (fig. 6), 

 but also between lots of fish taken by different 

 fishing methods, i. e., fish from different habitats 

 or depths within an area. To examine the in- 

 fluence of habitat on the composition of the food, 

 we may compare the same six lots of fish, using 

 the percentage of total volume as a measure of 

 importance for each food group (fig. 7). The 

 localities were treated separately, since we had 

 found that locality differences did exist. Figure 

 7 indicates that there was great variation in food 

 composition between similar habitats from dif- 

 ferent localities, and also for different habitats 

 within localities. The only apparent consistent 

 relationship is that the proportion of stomatopods 

 was higher in the subsurface-caught (longline) 

 fish than in the surface-caught (troll and pole-and- 

 line) fish. 



Figure 8 demonstrates the variation in avail- 

 ability of the major food groups in relation to 

 distance the tuna was captured from land. Since 

 fish taken at the surface came almost entirely 

 from inshore regions, the major data available 

 for this comparison are those from the longline- 

 caught fish. For both size groups of yellowfin, 

 1,299 mm. or less and 1,300 mm. or more, the aver- 

 age volume of Crustacea per stomach was much 

 greater for the fish taken near land (0 to 24 miles 

 from shore) than for those from offshore areas. 

 The crustacean fraction changed not only in its 

 average volume, but also in its composition; i. e., 

 for the near-shore region it was composed pre- 

 dominantly of crab and stomatopod larvae, 

 whereas for the offshore region it consisted of 



