412 



Fishery Bulletin 103(2) 



Results 



A total of 2761 hakes between 5 and 90 cm TL were 

 collected (Table 1). The total number of prey was about 

 1700, divided into 46 different species. Cluster and 

 NMDS analysis (stress = 0.02) based on the IRI allowed 

 the identification of four groups below 50% similarity 

 that were separated along a size gradient (Fig. 1). 



Euphausiids iNictiphan.es couchi, IRI=76%) and my- 

 sids (Lophogaster typicus, IRI=22%) dominated the diet 

 of group A (hake between 5 and 10.9 cm TL), and deca- 

 pods were the secondary prey. 



£ 40  



M 80 



100 



VI III IV V IX VII VIII 



Hake size classes 



Stress: 0.02 



Group A 



(<11 cm) II 



Group B (11 to15.9cm) 



Group C (16 to 35.9 cm) 



Figure 1 



Dendrogram and NMDS (nonmetric multidimensional scaling) 

 plot, based on IRI% values, of the nine hake (Merluccius merluc- 

 eius) size classes using group-average clustering from Bray-Curtis 

 similarity on diet data. (A) The four groups defined at arbitrary 

 similarity level of 50% are indicated (dotted line); (B) NMDS 

 showing the ordination of hake into four size classes with similar 

 diets (the details of each size class are explained in the text). 



Group B (hake from 11 to 15.9 cm TL) showed a more 

 heterogeneous diet characterized by a high occurrence 

 of euphausiids but also with a considerable number of 

 decapods (IRI=18%). Decapods were represented by a 

 wide variety of species, such as Chlorotocus crassicor- 

 nis, Alpheus glaber, Plesionika heterocarpus, Pasiphaea 

 sivado, and Solenocera membranacea. Pisces and mysids 

 showed lower percentages (IRI=15% and 4%, respec- 

 tively). Sepiolidae (IRI = 0.9%), Sepietta oweniana and 

 Alloteuthis media, dominated among cephalopods. 



The data suggest a gradual change towards a fully 

 piscivorous diet (Fig. 2) which begins around 16 cm TL 

 and is completed when sexual maturity is at- 

 tained (TL = 32 cm for males and TL = 38.5 cm 

 for females; Colloca et al., 2002). 



The importance of teleosts strongly increased 

 in group C (hake from 16 to 35.9 cm TL), where 

 they accounted for 91% of hake diet. The main 

 prey were Clupeiformes (IRI=61 %), Sardina 

 pilchardus and Engraulis encrasicolus. Fish 

 (IRI = 96%) represented almost the entire diet 

 of group D (>36 cm TL). In this group a shift 

 towards Centracanthidae (Spicara flexuosa, 

 Centracanthus cirrus) and a simultaneous de- 

 cline in consumption of Clupeiformes was ob- 

 served. Among decapods (IRI=4%), two species 

 occurred most frequently: Processa spp. and S. 

 membranacea. Euphausiids, mysids, and cepha- 

 lopods were absent in the diet of hakes larger 

 than 36 cm TL. 



Cannibalism of hake juveniles also accounted 

 for some of the diet and increased with predator 

 size. In hake between 36 and 40 cm TL cannibal- 

 ism represented 12% of IRI, reaching the highest 

 values (IRI = 17%) among larger individuals (TL 

 >51 cm). 



Discussion 



Hake is a top predator that occupies different 

 trophic levels during its ontogenetic develop- 

 ment. Hake size classes are differentiated along 

 food niche dimensions according to different prey 

 sizes or different prey taxa. Hake diet shifted 

 from euphausiids, consumed by the smaller 

 hakes (<16 cm TL), to fishes consumed by larger 

 hakes. Before the transition to the complete 

 icthyophagous phase, hake showed more gener- 

 alized feeding habits where decapods, benthic 

 (Gobiidae, Callionymus spp., Arnoglossus spp.) 

 and nectonic fish (S. pilchardus, E. encrasicolus) 

 dominated the diet, and cephalopods had a lower 

 incidence. Specific size-related differences in 

 prey spectrum seem to be associated with dif- 

 ferent spatial distributions or genetic needs (or 

 with both) (Flamigni, 1984; Jukic and Arneri, 

 1984; Velasco and Olaso, 1998). 



The patterns observed in the present study 

 indicated a strong partitioning among hake 



