256 
Fishery Bulletin 115(2) 
Table 2 
Fork lengths, measured in centimeters, of yellowfin tuna (Thunnus albacares) and skipjack tuna {Katsuwonus pelamis) 
caught in mixed schools in the eastern tropical Pacific Ocean in 2005, with standard deviations (SDs) given in parentheses 
after the means. No.=number of samples. 
Zone 1 
Zone 2 
Zone 3 
No. 
Min 
Max 
Mean (SD) 
No. 
Min 
Max 
Mean (SD) 
No. 
Min 
Max 
Mean (SD) 
Yellowfin tuna 
124 
39.9 
112.1 
69.8 (12.8) 
129 
47.0 
129.3 
80.8 (17.8) 
186 
48.9 
130.0 
79.5 (19.0) 
Skipjack tuna 
104 
39.2 
67.5 
52.5 (4.6) 
79 
41.0 
84.5 
57.1 (8.1) 
33 
40.8 
69.4 
50.8 (8.5) 
An analysis of similarities (ANOSIM) and the simi¬ 
larity percentage (SIMPER) analysis were carried out 
to evaluate the differences in diet within and between 
each tuna species and to establish the contribution 
of each prey item, respectively, with PRIMER, vers. 
6.1.6 (PRIMER-E, Auckland, New Zealand). For both 
analyses, we used permutation-randomization methods 
based on the Bray-Curtis measure of similarity, also 
termed “the percentage difference,” which is related to 
the mean character difference (Clarke and Warwick, 
2001). The Bray-Curtis method operates at the spe¬ 
cies level, and therefore the mean similarity between 
groups (e.g., between males and females) can be ob¬ 
tained for each tuna species. 
Results from ANOSIM are represented on an R 
scale from -i-l to -1. An R value of -i-l indicates that 
all the most similar samples are within the same 
groups. An R value of 0 occurs if the high and low 
similarities are perfectly mixed and bear no relation¬ 
ship to the group and indicate a completely random 
distribution within the group. An R value of -1 indi¬ 
cates that the most similar samples are all outside of 
the groups. Values from ANOSIM between 0.2 and 1 
with significance set at P<0.05 indicate that the null 
hypothesis can be rejected and the tuna species would 
have significantly different diets (Clarke and Warwick, 
2001). The SIMPER analysis breaks down the percent¬ 
age contribution of each species to the observed simi¬ 
larity (or dissimilarity) between samples. Diet compar¬ 
isons between yellowfin tuna and skipjack tuna were 
evaluated by comparing individuals captured in the 
same set within each area. 
Results 
Cumulative curves for size classes and prey species 
We sampled 439 yellowfin tuna and 216 skipjack tuna 
from mixed schools. Yellowfin tuna were the larg¬ 
est (mean FL) in zone 2 and the smallest in zone 1. 
Skipjack tuna were similar in mean FL in all 3 zones 
(Table 2). For all the yellowfin tuna samples, 381 stom¬ 
achs (87%) contained food, and 58 (13%) were empty; 
whereas, for all the skipjack tuna samples, 109 stom¬ 
achs (51%) contained food, and 107 (49%) were empty 
(Table 3). 
Out of a total of 25 sets, yellowfin tuna occurred in 
all sets and skipjack tuna occurred in only 8 sets. For 
yellowfin tuna, 124 stomachs with food were obtained 
from 9 sets in zone 1, 91 were obtained from 7 sets 
in zone 2, and 166 were obtained from 9 sets in zone 
3. For skipjack tuna, 75 stomachs with food were ob¬ 
tained from 5 sets in zone 1, 22 were obtained from 
2 sets in zone 2, and 12 were obtained from 1 set in 
zone 3. Cumulative curves for the prey species for each 
zone showed that a sufficient number of stomachs were 
analyzed to characterize the diet of both species (Fig. 
2), with the CVs <0.05 in all areas. 
Digestive state of prey species 
The state of digestion of the prey of both yellowfin and 
skipjack tunas varied widely in each sample zone. In 
zone 1, both tuna species had the highest number of 
prey species in digestive level 2. In zone 2, prey species 
in digestion level 4 were the most common whereas in 
zone 3, prey species were most often at digestive level 
3 (Table 4). 
Diet composition in zone 1 
Yellowfin tuna prey items comprised 21 taxa (9 cepha- 
lopods, 2 crustaceans, and 10 fish species). The %MN 
and %MW indices indicated that the most important 
prey items were the pelagic red crab (48.9%; 54.2%), 
the jumbo squid {Dosidicus gigas) (20.7%; 16.6%), and 
the bigeye cigarfish iCubiceps pauciradiatus) (13.4%; 
13.9%), respectively. According to the IRI, pelagic red 
crab (61.1%) and jumbo squid (28.7%) were the most 
important components in the diet (Table 5). The diver¬ 
sity index was 1.4 and R; was 0.1. The trophic level of 
the stomach contents was estimated at 3.9 (SD 0.4). 
The ANOSIM test indicated a similar diet composition 
by sex (i?=0.009), and differences among size classes 
(72=0.300, P=0.04; small individuals primarily con¬ 
sumed jumbo squid and pelagic red crabs; whereas 
large individuals primarily consumed Cubiceps spp.), 
months (72=0.360, P=0.01; in February individuals 
primarily consumed Cubiceps spp., jumbo squid, and 
