areas. The 15 species of fish were not omnipresent and 

 the data were only included in the calculations if 10 or 

 more individuals of that particular species were collect- 

 ed in a geographic area for stomach contents analysis. 



Only 6 of the 15 predators were collected in suf- 

 ficiently large numbers in the Middle Atlantic to war- 

 rant an analysis of dietary overlap. Three of these six 

 fish — silver, red, and spotted hake — showed inter- 

 mediate levels of dietary overlap, while the similarity in 

 diet for the haddock, fawn cusk-eel, and ocean pout was 

 low. 



In Southern New England 10 of the 15 predators were 

 present. A high degree of similarity in diet occurred be- 

 tween the red and spotted hake. These two fish fed on 

 both fish and invertebrates, and shared many of the 

 same species as prey (Tables 8, 9). Intermediate levels of 

 dietar>' overlap were generally found between the four 

 piscivorous predators, Atlantic cod, silver hake, white 

 hake, and offshore hake, and also between these same 

 four fish and the red and spotted hake. The only other 

 intermediate levels of overlap occurred between the fawn 

 cusk-eel and the red hake and haddock. Little similarity 

 was found between the diet of the haddock, longfin hake, 

 fawn cusk-eel, ocean pwut, and the other more pisci- 

 vorous predators mentioned above. 



For the fish taken on Georges Bank high levels of 

 dietary overlap occurred between the offshore and silver 

 hake and between the pollock and longfin hake. The high 

 overlap between the pollock, which have normally been 

 considered piscivorous, and the longfin hake, an inver- 

 tebrate feeder, may be explained by the dependence on 

 Crustacea as a major food source for Georges Bank pol- 

 lock (Table 3). It should also be noted that the average 

 length of pollock on Georges Bank was smaller than in 

 the other geographic areas which might explain the 

 reliance on crustaceans rather than fish. Intermediate 

 levels of diet similarity were found for a number of fish. 

 Most notable is the overlap between Atlantic cod and 

 haddock which may be traced to the heavy predation on 

 herring eggs by the Georges Bank haddock. Since Atlan- 

 tic cod were preying on either juvenile or adult herring, 

 this measure of overlap is somewhat misleading. Low 

 levels of similarity were again observed between the 

 invertebrate feeders. 



In the Gulf of Maine high and intermediate levels of 

 dietary overlap were found between the Atlantic cod, 

 pollock, silver hake, white hake, and red hake, but little 

 similarity was observed between the diet of these same 

 predators and the haddock, longfin hake, fourbeard 

 rockling, and ocean pout. The cusk showed an inter- 

 mediate level of dietary overlap with the pollock, red 

 hake, and fourbeard rockling while the diet of the ocean 

 pout was found to have an intermediate degree of simi- 

 larity with the haddock. All the remaining fish showed a 

 low level of dietary overlap. 



In Western Nova Scotia the highest level of dietary 

 overlap was found between the Atlantic cod and white 

 hake. Intermediate overlap was generally found between 

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

 hake, and cusk. Longfm hake also showed an inter- 



mediate level of overlap with the pollock, silver hake, 

 and marlin-spike. The last intermediate level occurred 

 between the marlin-spike and haddock. Little similarity 

 in diet was found for the other gadiform fish collected on 

 the Scotian Shelf. 



DISCUSSION 



Food 



Atlantic cod. — Data on the diet of the Atlantic cod 

 occurring off the U.S. coast in the Gulf of Maine have 

 been summarized by Bigelow and Schroeder (1953) (see 

 also Wise 1961), but, surprisingly, there is little other 

 information for the region of the present study (e.g., Ken- 

 dall 1898; Tyler 1972). The Canadian cod stocks in the 

 Gulf of St. Lawrence and off Newfoundland have, how- 

 ever, been more thoroughly investigated (Powles 1958; 

 Popova 1962; Templeman 1965; Kohler and Fitzgerald 

 1969; Minet and Perodou 1978). 



From some of the earliest studies it was concluded that 

 Atlantic cod will eat almost any food available. Despite a 

 wide diversity in the species composition of the diet, a 

 rather simple picture of the cod's food habits may be 

 drawn from most of the available data. In general, crus- 

 taceans and fish are the major prey, while the actual 

 percentage of the diet made up of each of these two tax- 

 onomic groups is determined by prey size and prey avail- 

 ability. Small cod prey on small crustaceans, such as am- 

 phipods and mysids. Slightly larger cod prey more on 

 decapod crabs and shrimp. The larger cod prey almost 

 exclusively on fish. Our observations on the food of 

 Atlantic cod caught off the U.S. coast are in agreement 

 with this rather simplistic description. Although we did 

 not separate the data into size classes, it can be seen that 

 fish and crustaceans were the most important prey cate- 

 gories (Table 1). We did not include the smaller cod (<20 

 cm) in our data base, and the average length of the 1,541 

 animals we analyzed was 54.7 cm (Table 1). Cod of 

 approximately 50 cm FL (fork length) have previously 

 been shown to prey heavily on both fish and crustaceans 

 (Powles 1958; Popova 1962; Rae 1968) so that our data 

 may be taken as confirmation that the diet of the U.S. 

 cod stocks is very similar to the other populations which 

 have previously been investigated. 



Pollock.— The diet of pollock was equally divided be- 

 tween fish and crustaceans (Table 1). Atlantic herring 

 was the major species of fish preyed upon while the 

 euphausiid, Meganyctiphanes norccgica, was the single 

 most important prey. The significance of euphausiids in 

 the diet of pollock had previously been noted by Kendall 

 (1898) who found that Thysanopoda were virtually the 

 exclusive prey for pollock taken off Eastport, Maine. Big- 

 elow and Schroeder (1953) included Kendall's observa- 

 tions in their treatise on fishes in the Gulf of Maine but 

 also mentioned that young herring were an equally im- 

 portant prey. Steele (1963) and Dexter (1969) again con- 

 firmed the importance of euphausiids such as M. nor- 

 vegica as major prey for pollock in the Gulf of Maine. 



18 



