Garcia-Rodriguez et al.: Ontogenetic shifts and feeding strategies of Gadiformes in the Mediterranean Sea 53 
Table 1 
Summary information related to the 7 most abundant species of Gadiformes sampled in the 
western Mediterranean Sea between 2011 and 2017. For each species, the scientific and common 
names, depth range (in meters), number of stomachs for which contents were examined, and size 
range (total lengths in millimeters) are provided. 
Scientific name Common name 
Gadiculus argenteus Silvery pout 
Gaidropsarus biscayensis 
Molva macrophthalma 
Merluccius merluccius 
Micromesistius poutassou 
Phycis blennoides 
Trisopterus minutus 
European hake 
Blue whiting 
Poor cod 
Dietary indices 
Three conventional dietary indices were calculated to 
provide quantitative information on diet compositions of 
the gadiform species investigated. A raw data set includ- 
ing prey species for each predator species is included in 
Supplementary Table 1 (online only). The first index used 
was frequency of occurrence (%FO), defined as the num- 
ber of stomachs containing a prey item compared with 
the total number of stomachs examined. The second 
index was the number index (%N), defined as the num- 
ber of individuals of a prey item compared with the total 
number of individual prey ingested. The third index cal- 
culated was the volume index (%V), which represents the 
volume of a prey item compared with the total volume of 
ingested prey (Hyslop, 1980). Finally, the geometric index 
of importance (%GII), which incorporates both %FO and 
%V, was used to avoid redundancy in the combination 
of mathematically dependent measures instead of other 
traditional indices (Tirasin and Jgrgensen, 1999). It was 
computed as follows (Assis, 1996): 
%GII, = aaa /n, (1) 
J 
where V;, = the value of the ith relative measure of prey 
quantity for the prey group j (in this case, 
Vj;=%FO+%V), and 
n =the number of the relative measures of prey 
quantity used in the analysis (in this case, n=2, 
%FO and %V). 
Therefore, in this study, the %GII was calculated as 
(%V+%FO)/2. Values for all indices are given as percentages. 
Ontogenetic shifts in diet 
Before identifying when ontogenetic shifts occur and 
establishing trophic groups based on different size 
classes, 10 major prey groups were identified according to 
Mediterranean bigeye rockling 
Mediterranean ling 
Greater forkbeard 
Depth No. of Size 
range stomachs range 
109-647 263 
59-634 116 
86-634 164 
34-722 1254 
76-722 786 
53-728 286 
36-352 323 
60-133 
40-180 
83-740 
87-549 
89-377 
85-393 
78-272 
taxonomic criteria and, in the case of fish species, habitat 
type criteria: small plankton; large plankton; Polychaeta; 
Cephalopoda; crab, lobster, and mantis shrimp species; 
shrimp species; Peracarida; demersal fish species; pelagic 
fish species; and benthopelagic fish species. 
To investigate variation in a predator’s diet according 
to size, trends in the volume index values of the most rep- 
resentative prey groups, mentioned in the previous para- 
graph, were plotted against fish length of predators to 
define the different categories that combine species and 
size class. We first compared both quantitative and qual- 
itative methods. Results indicate similar outcomes for 
some species, such as the European hake or blue whiting, 
and outcomes were less realistic for those species with a 
small number of samples. Therefore, we decided to use 
a qualitative method based on a graphic technique that 
allows easy identification of the size at which the trend 
in the trophic strategy changed for each species studied. 
In the quantitative method, the number of size classes 
was estimated following the Sturges procedure (Scherrer, 
1984), and in the qualitative method, the lengths were 
represented continuously at every millimeter along the 
x-axis of graphs. 
This graphic technique allows qualitative identification 
of the size at which a trophic shift occurred. We used this 
size for each gadiform species studied to divide size classes 
into /arge and small categories (sizes varied depending on 
the species; Table 2). In addition, size at first maturity was 
plotted with the aim of linking it with the size at which a 
trophic shift occurred. Sizes at first maturity for all the 
studied species were based on previously published data 
from studies in nearby areas (Biagi et al., 1992; Benghali 
et al., 2014; European Parliament and Council, 2019; 
Ismen et al., 2019). 
Trophic indices to describe feeding strategies 
To describe the degree of dietary diversity of a given spe- 
cies, we used niche breadth in accordance with Levins’s 
