FISHERY BULLETIN: VOL. 83, NO. 3 



The principal food item in the third quarter based 

 on both measurements was M. muelleri (Fig. 4). The 

 ranks for the other items did not correspond. Eu- 

 phausia similis was second in importance according 

 to the IRI, but ranked third next to other fish based 

 on the MVRM. Scomber japonicus and Benthodes- 

 mus sp. were the predominant species consumed in 

 the fish category. 



The IRI and MVRM ranks in the fourth quarter 

 corresponded with the exception of the principal 

 prey type (Fig. 4). Again, E. similis ranked first ac- 

 cording to the IRI, but Thyrsitops lepidopoides was 

 the primary food item based on the MVRM. Cepha- 

 lopods, mainly Argonauta sp., were consumed in 

 significant proportions in this quarter {IRI = 31.7, 

 MVRM = 7.2%). 



In summary, M. muelleri and E. similis predom- 

 inated in the skipjack tuna diet during all quarters 

 (Fig. 3). With the exception of the second quarter, S. 

 japonicus was an important food item. Benthodesmus 

 sp., C. ruber, A. thazard, as well as the cephalopod, 

 Argonauta sp., also proved important in specific 

 quarters. The importance of T. lepidopoides in 

 Figure 3 was exaggerated by its predominance in the 

 fourth quarter. 



Variations with Size 



As might be expected, basic dietary changes occur 

 as the skipjack tuna grow. Nakamura (1965), Alver- 

 son (1963), Batts (1972), Dragovich and Potthoff 

 (1972), and Wilson (1982) observed a decrease in the 

 relative importance of crustaceans and an increasing 

 importance of fish in the diet, as the skipjack size in- 

 creased. 



To evaluate the relationship between size and food 

 habits, the skipjack were arbitrarily divided into 

 seven 5 cm groups (Fig. 2). For each prey category 

 the MVRMs were stratified by size group and 

 (quarter (App. Tables 6-10, Fig. 5). Trends reported 

 in the results for length groups > 70 cm may not 

 represent feeding habits of skipjack tuna from the 

 study area because the sample sizes were too small 

 (Fig. 2. App. Table 6). 



There were no significant differences (P < 0.05) in 

 diet and size in the first quarter except in the amount 

 of other fish consumed (Fig. 5, App. Table 7). The 

 MVRM of this category increased from 28.7% in 

 skipjack 45.0-49.9 cm to 91.7% in fish 75.0-79.9 cm. 



In the second quarter the proportions of other fish, 

 other crustaceans, E. similis, and M. muelleri signi- 

 ficantly changed with size (Fig. 5, App. Table 8). The 

 larger skipjack tuna ate more fish {MVRM = 1 .7% in 

 the 45.0-49.9 cm size class toMVRM = 54.1% in the 



75.0-79.9 cm size class) and more M. muelleri 

 {MVRM = 15.4% to MVRM = 30.3%) than the 

 smaller skipjack. As their size increased, skipjack 

 decreased their consumption of E. similis from 

 MVRM = 76.9% to MVRM = 4.5%. There was a 

 significant difference between size classes {P < 0.05) 

 in the MVRM of other crustaceans in the diet, but 

 this difference seemed uncorrected to increases in 

 size. 



In the third quarter there were no significant dif- 

 ferences in the diet with increasing size except that 

 the MVRM of E. similis decreased from 25.0% to 

 8.0% (Fig. 5, App. Table 9). 



Thyrsitops lepidopoides was eaten by the smaller 

 skipjack (45-59.9 cm) only in the fourth quarter (Fig. 

 5, App. Table 10). Although there were significant 

 differences {P < 0.05) in the diets between the seven 

 size groups during this period, these differences 

 again seemed unrelated to increasing size. 



In summary, when the data on T. lepidopoides 

 were included with the rest of the fish data, there 

 were no significant differences between size groups 

 in the proportions of other fish consumed throughout 

 the year (Fig. 5, App. Table 6). The MVRM of E. 

 similis in the diet decreased from 42.5% in skipjack 

 tuna 45.0-49.9 cm to 0.0% in skipjack 75.0-79.9 cm. 

 There were significant differences in the percent- 

 ages of M. muelleri and S. japoniciis between the 

 size classes. There were no significant differences in 

 the MVRM of other crustaceans with changes in size. 



As reported in the studies of the food habits of 

 skipjack tuna referred to above, the stomach con- 

 tents of skipjack from this area indicated the basic 

 dietary changes associated with increasing size: a 

 significant decrease in the proportion of E. similis, 

 the predominant crustacean prey, and an increase in 

 proportion of M. muelleri, the predominant fish 

 prey. 



DISCUSSION 



Several studies have reported that skipjack tuna 

 feed predominantly on euphausiids and gonostoma- 

 tids. Dragovich and Potthoff (1972) reported that the 

 gonostomatid Vinciguerrin nimbario contributed 

 44.7% by volume to the diet of skipjack tuna from 

 the Gulf of Guinea. Zavala-Camin (1981) reported M. 

 muelleri and euphausiids as dominant food items in 

 36 stomachs of skipjack caught off Brazil. Alverson 

 (1963) reported skipjack tuna captured in the eastern 

 tropical Pacific fed primarily on euphausiids (47% by 

 volume, in 37% of the stomachs), followed in impor- 

 tance by the gonostomatid Vinciguerria lucetia 

 (10% by volume). The abundance of euphausiids in 



384 



