ophiuroids (9.6%), such as Ophiopholis (1.1%), and the 

 pelecypod, Pecten (4.5%), formed the remaining bulk of 

 the diet. In the Middle Atlantic few ocean pout were col- 

 lected, but as on Georges Bank and in Southern New 

 England, the single major prey species was the sand dol- 

 lar (86.3%). In the Gulf of Maine echinoderms were 

 apparently less important although E. parma still ac- 

 counted for 24.8'; of the prey. "Other Groups" was also 

 important and the single contributor to this category was 

 the tunicate, Cnemidocarpa mollis (18.5'o). Sand and 

 rock (33.4%) made up a large part of the weight of the 

 stomach contents. 



Dietary Overlap 



The percentage similarity values (Fig. 3) calculated 

 from the data in Table 2 range from 1 to a maximum of 

 75. In order to facilitate the comparison between species, 

 similarity values have been grouped into three cate- 

 gories representing low (0 to <30%), intermediate (30 to 

 <60%), and high (60 to 100%) levels of dietary overlap. 

 Accordingly, it is obvious that the greatest similarity in 

 diet exists between the silver and white hake (75%), the 

 marlin-spike and longnose grenadier (75%), the red and 



spotted hake (71%), and the pollock and silver hake 

 (63%). 



The diets of Atlantic cod, pollock, silver hake, white 

 hake, offshore hake, cusk, red hake, and spotted hake are 

 reasonably similar. Most of these fish populations are 

 primarily piscivorous, which accounts for the observed 

 intermediate and high levels of dietary overlap. The red 

 and spotted hake, however, qualify as "mixed feeders" 

 preying on both fish and invertebrates. In contrast to the 

 eight predators listed above, the remaining seven 

 species: haddock, longfin hake, fourbeard rockling, mar- 

 lin-spike, longnose grenadier, fawn cusk-eel, and ocean 

 pout, prey almost exclusively on invertebrates. Of these 

 fish only the haddock preyed extensively on other fish 

 and this was predation on herring eggs rather than on 

 juvenile or adult fishes. The lack of similarity between 

 the diet of the invertebrate feeders is in part due to the 

 more numerous categories of potential prey. In some in- 

 stances, such as for the marlin-spike, longnose gren- 

 adier, and fourbeard rockling, estimates of diet similarly 

 are of limited value because of the small number of fish 

 examined and/or the small amount of prey found in the 

 stomachs. 



The percentage similarity in diet has also been cal- 

 culated for each fish in the five individual geographic 



30 

 < 60 



Figure 3. — Percentage similarity between llic diet of 15 species of gadiform fishes from the northwest .Atlantic. For calculation methods, 

 seetext. Open boxes = to 00%; ruled boxes = .10 to <607c; solid boxes = 60 to 100%. 



17 



