456 
Fishery Bulletin 115(4) 
Table 3 
Comparison of prey items in the diet of smooth hammerhead (Sphyrna zygaena ) collected from 
December 2012 through June 2015 in northern Peru, generated from a Bray-Curtis index that 
is based on the percentage by number of prey (%N) and from R-statistics and P-values gener¬ 
ated from analysis of similarities. Size classes of sharks: I (53-70-cm-TL), II (71-100 cm TL), 
III (101-190 cm TL) and IV (191-294 cm TL). An asterisk (*) indicates comparisons for which 
differences were significant (P<0.001). “Overlapping” indicates that the diet of smooth hammer¬ 
head overlaps for the 2 size classes in each paired comparison (of horizontal and vertical values). 
Size class I Size class II Size class III 
Overlapping 
R 
Overlapping 
R 
Overlapping 
R 
Size class II 
33 
0.47* 
_ 
_ 
_ 
_ 
Size class III 
25 
0.71* 
66 
0.27* 
— 
— 
Size class IV 
36 
0.29* 
26 
0.6* 
32 
0.2 
similar to the result of 4.3 that we found in this study 
(Castaneda, 2004; Estupinan-Montano and Cedeno- 
Figueroa, 2005; Bolano Martinez, 2009; Galvan-Maga- 
na et al., 2013). Additionally, the fact that 92% of 
stomach contents examined were in an advanced state 
of digestion indicates that smooth hammerhead is an 
intermittent feeder. Stomach contents of a continuous 
feeder would have food items at different stages of di¬ 
gestion (Medved et ah, 1985). 
Although we were able to analyze the contribution 
of cephalopods in the diet, we were limited in our abil¬ 
ity to quantify the contribution of fish in the diet. Ac¬ 
cording to the percentage of prey by frequency of oc¬ 
currence, 12.9% of the diet was composed of fish that 
could not be identified at a species level owing to their 
advance state of digestion. Otoliths are often used 
to identify species because they resist the digestive 
process. However, in our study, smooth hammerhead 
preyed upon pelagic fishes with small and expellable 
otoliths, often preventing species identification (Lom- 
barte et al., 2010). Conversely, the hard structures of 
squid beaks were easier to detect owing to their larger 
sizes and resistance to digestion (Braccini et al., 2005). 
Despite these challenges, although fish as prey may be 
underestimated, we were still able to identify otoliths 
and it is clear that fish comprise an important com¬ 
ponent of the diet of the smooth hammerhead. Future 
studies could better refine these estimates with the use 
of complementary methods, such as molecular analysis, 
that are helpful in identifying taxonomic groups with 
precision (King et al., 2008). Moreover, future studies 
should emphasize the collection of samples from sharks 
greater than 200 cm TL. We were able to collect only 
10 samples for size class IV (191-294 cm TL). There¬ 
fore our results more accurately represent the diet of 
neonates and juveniles. 
Life stages 
Shark species change their diet over the course of 
their life (Lowe et al., 1996; Wetherbee and Cortes, 
2004). In Ecuador, as smooth hammerhead grew, Pa¬ 
tagonian squid decreased proportionally in the over¬ 
all diet composition; whereas jumbo flying squid in¬ 
creased (Bolano Martinez, 2009). Similarly, in South 
Africa, juveniles fed on loliginids and adults fed on 
Ancistrocheirus sp. and red flying squid ( Ommas - 
trephes bartramii) (Smale and Cliff, 1998). Our results 
are consistent with those reported for Ecuador and 
South Africa. We found that all size classes had sta¬ 
tistically significant differences, except for size classes 
III and IV, which could be explained by the low sam¬ 
ple size of size class IV. Moreover, our samples were 
composed mainly of neonates and juveniles and in¬ 
cluded only a small sample size of adults. Therefore, 
further studies, should include a wider range of sizes 
to assess fully the trophic ecology of smooth hammer¬ 
head over its entire size range. 
We found that neonates and small juveniles con¬ 
sumed coastal species (i.e., Patagonian squid, Peprilus 
sp.; Jereb and Roper, 2010), and larger juveniles and 
adults consumed oceanic species (i.e., jumbo squid, An¬ 
cistrocheirus lesueurii ; Nigmatullin et al., 2001; Jereb 
and Roper, 2010). These diet habits suggest a change 
of habitat and distribution. Sharks of size class IV 
were the only individuals that consumed giant squid, 
which is a deep-sea species with a vertical distribution 
range of 200-1000 m (Landman et al., 2004; Jereb and 
Roper, 2010). In New Zealand, an electronically tagged 
smooth hammerhead measuring 160 cm TL gave evi¬ 
dence of vertical migrations and a maximum depth re¬ 
corded at 144 m (Francis, 2016). This finding suggests 
that larger sharks may be migrating vertically to cap¬ 
ture prey. Furthermore, the change in diet from coastal 
to oceanic prey species can be explained partly by the 
need to consume prey species of greater biomass and 
energy content (Navia et al., 2007). The Patagonian 
squid, for example, provides 3.1 kJ/g, whereas jumbo 
flying squid provides 6.6 kJ/g (Croxall and Prince, 
1982; Abitia-Cardenas et al., 1997). Trophic position of 
this species rises to a higher level in the food chain as 
the sharks increased in size, and this has also been ob- 
