Page 1 



INTRODUCTION 



Trophic structure offish assemblages on the continen- 

 tal shelf from New England to Cape Hatteras, North Caro- 

 lina, has been examined by the Northeast Fisheries Science 

 Center (NEFSC) in several large-scale dietary studies. Past 

 studies have described food-web structure and trophic in- 

 teractions among Northwest Atlantic fishes. For the 1969- 

 72 study period, see Maurer and Bowman (1975) and 

 Langton and Bowman ( 1 980, 1 98 1 ); for the 1 973-76 period, 

 see Edwards and Bowman ( 1 979) and Bowman and Michaels 

 ( 1 984). Also see Cohen et al. ( 1 982), Sissenwine ( 1 984), and 

 Sherman (1986). 



Understanding trophic interrelationships among the 

 majority of fish species within an ecosystem is necessary 

 to define more precisely the role that predation plays in 

 determining ecosystem structure and the possible long-term 

 effects of various fisheries exploitation regimes. The pri- 

 mary purpose of this report is to provide basic diet compo- 

 sition data on fishes and two species of squids commonly 

 caught in the Northeast Continental Shelf Ecosystem. 



Diet data for some of the species covered in this report 

 are scant or nonexistent in the published literature. This 

 report's data expand on existing diet data for major fish and 

 squid species sampled during the 1 969-72 and 1 973-76 study 

 periods, and cover the broader geographic area from Nova 

 Scotia to Cape Fear, North Carolina, including inshore areas 

 where bottom depth ranges from 8 to 27 m. Previous stud- 

 ies only covered the area from Nova Scotia to Cape Hatteras 

 in water depths of 27-366 m. The majority of data presented 

 here represent the last of a series of quantitative stomach 

 content collections initiated in 1973. 



Detailed stomach content data, based on percentage 

 composition by weight or volume, are presented for indi- 

 vidual predator species. Also, we identify six major func- 

 tional prey groups based on the predators' stomach con- 

 tents. 



METHODS AND MATERIALS 



Specimens sampled for stomach content analysis were 

 primarily collected during NEFSC bottom trawl survey 

 cruises conducted during the spring, summer, autumn, and 

 winter from 1977 to 1980. Stratified random sampling was 

 conducted in continental shelf waters from Nova Scotia to 

 Cape Fear, North Carolina and sampling occurred 24 hr/day. 

 Tows were 30 min in duration at a vessel speed of 6.5 km/hr, 

 usually in the direction of the next sampling station. Bot- 

 tom depths sampled ranged from 8 to 366 m. The 27-m 

 depth contour (along the coast) delineates inshore versus 

 offshore sampling areas of the NEFSC. Eight general areas 

 surveyed by the NEFSC are depicted in Figure 1 . They 

 include the six traditional offshore areas - offshore south 

 of Cape Hatteras, Middle Atlantic, Southern New England, 

 Georges Bank, Gulf of Maine, and Scotian Shelf-- along 



with two inshore areas ~ inshore south of Cape Hatteras 

 and inshore north of Cape Hatteras. 



Stomach content samples taken during bottom trawl 

 survey cruises were based on three criteria as follows: 1) 

 offshore species of particular interest to investigators at 

 the NEFSC for a variety of reasons (e.g., species making up 

 the majority of commercial catches or species known to 

 consume primarily fish), 2) species making up the majority 

 of catches (by weight) in inshore areas (hitherto not 

 sampled), and 3) species for which dietary information was 

 scant or nonexistent. Survey technicians sampled species 

 first for criterion no. 1 if the station was offshore, or for 

 criterion no. 2 if the station was inshore. Other samples 

 were taken when time permitted. Samples generally repre- 

 sented the length frequency of each species caught. 



Stomachs of large fish or squid were excised aboard 

 ship, individually wrapped in gauze with a label denoting 

 vessel, cruise, station, species, specimen size (i.e., fork length 

 when applicable, otherwise total length, disk width for rays, 

 or mantle length for squid), sex, and maturity, and preserved 

 in a 3.7% formaldehyde solution (Formalin and sea water) 

 by volume. Small fish and squid were preserved whole. 



The preserved stomachs were individually opened in 

 the laboratory and their contents emptied onto a 0.25-mm- 

 mesh-opening screen sieve to permit washing away the form- 

 aldehyde without the loss of any food items. The stomach 

 contents were sorted, identified, counted, and damp dried 

 on absorbent paper. Major prey items and commonly oc- 

 curring but relatively minor prey, in terms of weight, were 

 identified to species whenever practical. The wet weight of 

 all stomach content groups was determined to the nearest 

 0.001 g, and all data recorded. A stomach was considered 

 empty when no material was found in the stomach, or when 

 the material found in the stomach both could not be identi- 

 fied and weighed less than 0.001 g. 



We also provide information on the food of large pe- 

 lagic species. These samples were gathered from various 

 sources during 1963-84. Stomach content samples of apex 

 predators, including large sharks and tunas, and other large 

 species (/'. e. , > 1 00 cm) were mostly collected from fish caught 

 by rod and reel, or by longline during research vessel cruises. 

 Some samples were collected from fish caught during fish- 

 ing tournaments over the years. The sampling area cov- 

 ered continental shelf waters from Florida to the Grand 

 Banks (southeast of Newfoundland). 



As noted earlier, stomach content data associated with 

 the 1 977-80 period of bottom trawl survey cruises were mea- 

 sured as percentage composition by wet weight. Data as- 

 sociated with the 1963-66 period were measured as percent- 

 age composition by occurrence. Data associated with the 

 1969-72 period are based on samples first being grouped 

 according to fish length, then being measured as percent- 

 age composition by wet weight. Stomach content data pre- 

 sented for apex predators are based on percentage compo- 

 sition by volume. 



No statistical weighting (e.g., length frequency, sample 

 size, population size, or species distribution) was applied to 



