FOOD OF ALBACORE IN THE PACIFIC 



463 



fishing method. Where albacore of approximately 

 the same size were considered, the average volume 

 of food per stomach was also used as a basis for 

 comparison. In some instances, weights of the 

 albacore were estimated from length to weight 

 tables developed at the Bureau of Commercial 

 Fisheries Biological Laboratory', Honolulu. 



Statistical tests of significance have not been 

 made, for, as King and Ikehara (1956) pointed 

 out: "Regardless of the mehods of analysis used, 

 there are many uncontrollable variables inherent 

 in food studies which detract from the precision 

 of the results." There is evidence (Reintjes and 

 King, 1953, fig. 4) that the parameters involved 

 are not independent, and therefore the assumptions 

 underlying the common tests of significance would 

 be violated. 



In a number of my comparisons of stomach con- 

 tent data with environmental variables, the 

 stomach data represent catches made during differ- 

 ent cmises and years. Little can be done to rectify 

 this weakness, since further grouping of the 

 stomach data into subclass numbers more, discrete 

 than those shown would produce subsamples of 

 very small numbers. 



RESULTS 



VARIATION IN FOOD WITH FISHING METHOD 

 AND SIZE OF THE ALBACORE 



The following discussion of variations in the 

 food of albacore with the method of capture by 

 inference is a generalization on variations in food 

 with albacore size, since the methods of capture 

 sampled different size groups (fig. 2) . 



There are distinct differences in the average 

 volume of food per stomach and in the composi- 

 tion of the foodstuffs dejwnding upon which fish- 

 ing method was used. Table 2 shows the average 

 volume per stomach, figiire 4 sliows the distribu- 

 tion of tliese volumes, and figure 5 shows the com- 

 parative importance, by volume, of the major food 

 groups of stomach contents, according to method 

 of capture. 



Table 2. — Avei-age stomach valunies of Sj^S albacore. ac- 

 cording to method of capture 



GILL NET (N=87) 



TROLL (N=79) 



LOHaHE. (N'I82) 



10 15 20 30 40 50 75 >IOO 

 149 199 299 399 499 749 999 



CC / STOMACH 



Figure 4. — Distribution of stomach content volumes or 

 348 albacore, according to method of capture. 



The higher average volume per stomach of the 

 longline-caught albacore is undoubtedly due to 

 their larger sizes, since the longline captured all 

 the albacore longer than 85 cm. Only 9 percent 

 of the longline-caught fish were under 85 cm. while 

 the majority of the gill net- and troll-caught al- 

 bacore were in the 50-70 cm. range. 



The difference in average stomach volume be- 

 tween the troll- and gill net-caught fish is not as 

 easily explained, since the fish of both groups 

 were approximately the same size. One possi- 

 bility is that this difference reflects the time of 

 feeding of albacore, because the troll-caught fish 

 are taken during the day and the gill netted alba- 

 core are thought to have been caught at night, 

 even though the gill net is hauled aboard after 

 dawn. An indication of this was provided by the 

 24-liour gill net station on John R. Manning 

 cruise 36. No albacore were caught by the sets 

 made from 0828 to 1531 hours and from 0230 to 

 0942 hours. Seventeen were caught by the set 

 from 2003 to 0358 hours, and 6 albacore were 

 caught in the set from 1502 to 2153 hours 



