Atlantic but contributed significantly less to the diet (1.9%). In this 

 region predation on silver hake was somewhat overshadowed by 

 predation on Gulf Stream flounder, Cilharichlhys arctifrons 

 (7.6"%). None of the other fish could be identified to the species 

 level. 



Molluscs were the third and final taxon that played an impor- 

 tant role in the diet of fourspot flounder. Within this phylum 

 the Cephalopoda accounted for most of the prey. In the Middle 

 Atlantic most of the cephalopods could not be identified below 

 the class level (9.8%), and the only genus identified was Rossia 

 (0.3%). On Georges Bank most of the caphalopods, again, were 

 not identified below class (10.3%) and here the one genus identi- 

 fied was Loligo (2.1%). In Southern New England three differ- 

 ent squid were identified in the flounder stomachs. Loligo, 

 Rossia [R. tenera (0.1%)], and ///ex illecebrosus accounted for 

 3.8, 2.6, and 2.8% of the diet, respectively. 



Windowpane. — The diet of windowpane can be compared in 

 three of the four areas where fish were collected for stomach 

 contents analysis. In all three areas, arthropods accounted for at 

 least three-fourths of the diet, and most of the prey within this 

 phylum were the more pelagic crustaceans (Table 4). The actual 

 composition of the diet, although very similar in terms of the 

 type of prey, varied quite appreciably in percentage composition. 



In the Middle Atlantic, mysids, especially Neomysis 

 americana (73.2%), were the primary prey. Crangon 

 septemspinosa (6.1%), Dichelopandalus leptocerus (4.2%), and 

 Caridion gordoni (2.2%) were of secondary importance. In 

 Southern New England, mysids, Neomysis americana (50.2%), 

 were again a major prey, but Dichelopandalus leptocerus 

 (31.9%) was also very important, and Crangon septemspinosa 

 accounted for approximately the same percentage of the diet 

 (3.3%) as in the Middle Atlantic. In contrast, on Georges Bank, 

 Crangon septemspinosa (45.9%) was the major prey, while 

 Neomysis americana (19.1%) and Dichelopandalus leptocerus 

 (6.0%) were of secondary importance. 



On Georges Bank, Pisces was the only taxon, besides the 

 Arthropoda, to contribute significantly to the diet (13.1%). 

 Four species of fish were identified from the Pisces remains: 

 American sand lance (Ammodytes americanus) larvae (5.1%) 

 and adults (4.4%); silverside, Menidia (0.8%); longhorn 

 sculpin, Myoxocephalus octodecimspinosus (0.7%); and finally 

 other windowpane (1.0%). The only other fish prey that was 

 identified to species was one blackbelly rosefish, Helicolenus 

 dactylopterus (0.7%), which was eaten by Southern New 

 England windowpane. 



Table 4.— Geographic breakdown of the stomach contents of windowpane 

 flounder, Scnphlhalmus aquosiis, in the northwest Atlantic. Data are expressed 

 as a percentage weight for fish collected during the spring and autumn bottom 

 trawl survey cruises, 1969-72. ( + indicates present but <0.I%.) 



Witch flounder. — Witch flounder were collected in all five 

 geographic areas, but the average total fish length varied by as 

 much as 24 cm between regions. Despite this range in length, the 

 composition of the witch floimder's diet is very similar. In all 

 regions, annelids were the major prey (Table 5). Annelids made up 

 more than half the diet in each area except Southern New England, 

 and here the Mollusca (40.4%) offset the importance of annelids 

 primarily because two specimens of the squid, Loligo, accounted 

 for 32.8% of the diet by weight. 



The total quantity of prey in the stomachs also puts some 

 limitations on detailed areal comparisons. In the Middle Atlantic, 

 for example, only 38 fish were collected which had consumed 

 slighdy > 10 g of prey. Although armelids accounted for 92.8% of 

 the diet, over 90% of these were unidentified polychaetes. Two 

 species were identified, Lumbrineris fragilis (<0.1%) and Eteone 



longa ( <0. 1 %). Lumbrineris fragilis was identified in the stomachs 

 from other areas, which shows that Middle Atlantic witch flounder 

 prey on at least one of the same species as in the other areas. 



A comparison of the annelid prey from the fish collected on 

 Georges Bank, in the Gulf of Maine, and from Western Nova 

 Scotia will be the best indicator of changes in feeding behavior, 

 since a reasonable numba of fish were collected in these areas, and 

 the average fish size was virtually the same. On Georges Bank, 

 40.3% of the annelid prey was unidentified, but, of the remaining 

 47.5%, Scalibregma ir\flatum (28.4%) was the most important. In 

 the other two areas, the Scalibregmidae were also preyed on, but 

 they accounted for <0.5% of the prey, e.g., in the Gulf of Maine, 



