currence for the food item was calculated by dividing 

 the number of stomachs that contained the item by 

 the total number of walleye pollock stomachs. 

 Stomachs with a trace of food were included in this 

 analysis, but empty stomachs were not. 



The volumetric data were categorized by size and 

 sex of walleye pollock and by season of year (for 

 dates, see Table 1) to determine if these factors in- 

 fluenced the diet. To compare types of food eaten by 

 different sizes of walleye pollock, the fish were arbi- 

 trarily divided into three length groups: Small, <250 

 mm; intermediate, 250-349 mm; and large, >349 mm 

 SL. The percent of empty stomachs was also 

 calculated for each season and size category as an in- 

 dicator of seasonal feeding activity. Because the size 

 of walleye pollock was generally different in each bay 

 during the same season, the foods of the fish in the 

 two bays could not be compared. Consequently, data 

 for the two bays were pooled for all analyses. Dif- 

 ferences in feeding at different depths also could not 

 be analyzed because, as previously noted, the precise 

 depth where fish were caught was unknown. 



Results and Discussion 



When data from all samples of walleye pollock were 

 combined (regardless of size, sex, or season), crus- 

 taceans were the major food. Of the crustaceans, 

 euphausiids and mysids had the highest percent 

 volumes and frequencies of occurrence (Tables 2, 3). 

 Thysanoessa raschii was the most frequently eaten 

 euphausiid; Acanthomysis pseudomacropsis and 



Neomysis kadiakensis were the most frequently 

 eaten mysids. The percent volume of shrimp was 

 nearly as high as the percent volume of euphausiids 

 and mysids; however, shrimp were eaten less fre- 

 quently. Most of the shrimp were either Crangon spp. 

 (Crangonidae) or Pandalidae. Copepods, hyperiid 

 and gammarid amphipods, and cumaceans were 

 found in many of the stomachs (> 20%), but their per- 

 cent volumes were small (<3%). Often stomachs con- 

 tained only one or two individuals that comprised a 

 minute fraction of the contents. Copepods were most- 

 ly Metridia sp. and Calanus sp.; cumaceans were 

 mostly Eudorella sp. Amphipods were generally not 

 identified to species. 



Fish and, to a lesser degree, polychaetes were the 

 only other foods present in amounts greater than a 

 trace. They were found less frequently than crus- 

 taceans and were usually identifiable only to class. 

 However, compared with crustaceans, fish found in 

 the stomachs were relatively large and usually com- 

 posed most of the volume of food in the stomachs, 

 when they occurred. 



The common food organisms in all of the studies of 

 walleye pollock have been crustaceans and fish. In 

 the eastern Bering Sea in summer 1970, walleye 

 pollock fed almost exclusively on euphausiids, 

 copepods, and fish (Takahashi and Yamaguchi 

 1972). In fall 1972, walleye pollock there ate eu- 

 phausiids and fish (Mito 1974); and in summer 1974 

 and spring 1977, they ate mostly euphausiids, 

 copepods, fish, and amphipods (Bailey and Dunn 

 1979). In southeastern Alaska in 1979-80 (my study), 



Table 2. — Mean percent volume of major categories of food in walleye pollock stomachs 

 from Auke Bay and Fritz Cove, southeastern Alaska, 1979-80. (Table does not include 

 empty stomachs or those with only a trace of food.) 





1 91 additional fish were examined for which sex could not be determined 



639 



