60 
Fishery Bulletin 119(1) 
Similarity 
Figure 3 
Multidimensional scaling ordination of stomach contents 
of specimens of 7 species of Gadiformes caught in the west- 
ern Mediterranean Sea from 2011 through 2017. In the 
(A) 2-dimensional and (B) 3-dimensional plots, dashed 
lines indicate or dotted lines surround predator groups 
(G1, G2, G3, and G4) identified by using cluster analysis 
based on feeding strategies. Results are given for 12 cat- 
egories that combine species and size class (small [S] and 
large [L]; length ranges for size classes are provided in 
Table 2). The horizontal dashed line in panel A indicates 
the threshold similarity for group selection. The species 
studied include the Mediterrenean ling (Molva macro- 
phtalma) (MOL), greater forkbeard (Phycis blennoides) 
(PHY), poor cod (Trisopterus minutus) (TRI), Mediterra- 
nean bigeye rockling (Gaidropsarus biscayensis) (GAID), 
silvery pout (Gadiculus argenteus) (GAD), blue whiting 
(Micromesistius poutassou) (MIC), and European hake 
(Merluccius merluccius) (MER). 
The last group was represented by both size classes of 
blue whiting and European hake. The small size classes of 
both of these species preyed mostly on large plankton, with 
euphausiids as the most prevalent prey (with %V values of 
51.38% and 24.3%, respectively). Benthopelagic fish species, 
namely the horned lanternfish (%V=19.3%) and the jewel 
lanternfish (Lampanyctus crocodilus) (%V=17.9%), were 
the main prey for the large size class of blue whiting. Pelagic 
fish species, such as the European anchovy (Engraulis 
encrasicolus) (%V=30.3%) and European pilchard (Sardina 
pilchardus) (%V=9.9%), were the species most widely con- 
sumed by the large size class of European hake. 
Co-occurrence and trophic overlap 
The highest percentages of co-occurrence were found for 
the associations of European hake with poor cod (75.8%) 
and blue whiting with greater forkbeard (64.1%) (Fig. 4). 
Conversely, the lowest values of co-occurrence were 
observed for Mediterranean ling with poor cod (11.3%) and 
European hake with Mediterranean ling (14.1%). These 
results, together with analysis of trophic levels, indicate a 
high niche overlap between silvery pout and blue whiting 
and a low degree of overlap between Mediterranean ling 
and poor cod. 
All species had an intraspecific trophic overlap greater 
than 0.50. Greater forkbeard had the lowest value (0.56), 
and silvery pout had the highest value (0.92). On the other 
hand, low co-occurrence was found for both blue whiting 
and Mediterranean ling (0.16 and 0.20, respectively), and 
high values of co-occurrence were found for European 
hake, greater forkbeard, and silvery pout (0.77, 0.69, and 
0.68, respectively). 
Discussion 
Our comprehensive study of feeding habits and trophic 
relationships among 7 fish species of the order Gadiformes 
that are important to fisheries in the western Mediterra- 
nean Sea explored ecological patterns such as those of 
ontogenetic variation in diet or trophic and spatial seg- 
regation of species. The results of this work can inform 
efforts to improve management of resource partitioning 
within and among these species in a broad geographic 
area (i.e., the entire GSA-06). 
Our data indicate that the species studied are struc- 
tured into 4 major guilds based on their feeding habits. 
The first group comprises the Mediterranean bigeye 
rockling, poor cod, and greater forkbeard. These species 
occupy medium-high positions within the food web. Their 
diets have previously been reported to consist of deca- 
pods and other small crustaceans, for example, for poor 
cod or greater forkbeard in the Catalan Sea (Macpher- 
son, 1978b), in the Gulf of Valencia (Morte et al., 2001, 
2002), and in the Adriatic Sea (Duléi¢ and Duléi¢, 2004). 
However, for Mediterranean bigeye rockling, our results 
differ from those of Macpherson (1978b), who identified 
Eusergestes arcticus and northern krill (Meganyctiph- 
anes norvegica) as the predominant prey. Our results 
indicate that Calocaris macandrae was the primary prey 
species of Mediterranean bigeye rockling, and it is worth 
noting that this species was also found to be cannibalis- 
tic. These differences may be related to the 38-year gap 
between the 2 studies or are possibly a result of the fact 
that our sampling did not explore seasonal dietary 
