314 
Fishery Bulletin 116(3-4) 
Table 1 ( continued) 
Prey species 
%FQ 
%N 
%PM 
%W 
%PW 
%PSIKI 
TELEOSTEI 
Clupeiformes 
Engraulidae 
Engraulis ringens 
2.61 
1.12 
42.91 
1.52 
58.43 
1.32 
Gadiformes 
Macrouridae 
Coryphaenoides sp. 
0.87 
0.02 
2.86 
<0.01 
<0.01 
0.01 
Merlucciidae 
Merluccius gayi peruanus 
0.87 
0.05 
6.25 
0.56 
63.94 
0.31 
Myctophiformes 
Myctophidae 
Myctophum aurolaternatum 
0.87 
0.09 
10.00 
<0.01 
0.03 
0.04 
Perciformes 
Coryphaenidae 
Coryphaena hippurus 
1.74 
0.43 
25.00 
1.65 
94.84 
1.04 
Scombridae 
Auxis thazard 
0.87 
0.08 
9.09 
0.02 
2.87 
0.05 
Sphyraenidae 
Sphyraena idiastes 
0.87 
0.11 
12.50 
<0.01 
<0.01 
0.05 
Unidentified fishes 
22.61 
8.71 
38.53 
3.51 
15.51 
6.11 
MAMMALIA 
Cetacea 
Delphinidae 
1.74 
0.23 
12.96 
0.91 
52.29 
0.57 
Delphinus capensis 
0.87 
0.29 
33.33 
0.25 
28.61 
0.27 
CRUSTACEA 
Decapoda 
Galatheidae 
Pleuroncodes monodon 
0.87 
0.84 
97.14 
0.81 
92.64 
0.83 
Unidentified crustaceans 
5.22 
0.70 
13.42 
<0.01 
0.06 
0.35 
AVES 
Unidentified bird 
0.87 
0.11 
12.50 
<0.01 
0.31 
0.06 
EGGS 
Exocoetidae eggs 
7.83 
2.17 
27.76 
2.25 
28.75 
2.21 
1993). The 12-statistic values from ANOSIM describe 
the extent of similarity, where R- 0 refers to an identi¬ 
cal diet and R- ±1 is indicative of the most divergent 
diet (Clarke, 1993). Similarity percentages (SIMPER) 
were used, as well, to determine the prey categories, 
by family, that typified particular groups, or contrib¬ 
uted most to the similarities between groups, or both 
(Clarke, 1993). Finally, the semiparametric permuta¬ 
tion multivariate analyses of variance (PERMANOVA) 
test on the Rray-Curtis index of dissimilarity (based 
on numeric abundance) was used to test whether an 
interaction was present between factors that were sig¬ 
nificant according to the pairwise tests (ANOSIM). 
Statistical analyses were conducted with the soft¬ 
ware RStudio, vers. 1.1.453 (RStudio, 2018) with R, 
vers. 3.5.0 (R Core Team, 2018). Descriptive statistics 
are presented with mean standard deviations (SDs). 
Results 
A total of 143 stomachs of blue sharks were obtained 
from the ports of Mancora (n=43) and Salaverry (n=100). 
Of these, 115 (80.4%) had food remains and 28 (19.6%) 
were empty. Of those individuals with food remains, 
47 were female (range: 165-293 cm TL; mean: 214 cm 
TL [SD 35] and 68 were male (range: 110-288 cm TL; 
mean: 220 cm TL [SD 41]). A cumulative prey curve was 
constructed on the basis of 74 stomach contents because 
41 stomachs containing only ‘unidentified remains’ were 
excluded. The curve slope was 0.9, indicating that the 
sampling size was sufficient to describe the diet of blue 
sharks. However, according to the Bizzarro’s et al. (2007) 
method, the curve did not reach an asymptote (P<G.05), 
and therefore suggests that the results presented here 
do not fully describe the blue shark diet. 
Prey comprised 42 taxonomic levels (Table 1). Cepha- 
lopods represented the main prey group (87.4% PSIRI), 
followed by fishes (8.5% PSIRI), flying fish eggs (2.1% 
PSIRI), crustaceans (1.1% PSIRI), and marine mammal 
and bird remains (0.8% and 0.1% PSIRI, respectively). 
The prey item ‘unidentified cephalopoda' (represented 
by flesh, gladii, and eye lenses) was the most impor¬ 
tant (31.3% PSIRI) in the overall analysis, followed by 
Argonauta spp. (16.1% PSIRI) and A. lesueurii (6.2% 
PSIRI) (Table 1; Fig. 2). 
