Table 4. — Grams of food organisms per fish (not including fish with empty stomachs) in 

 each size class, Oregon, summer 1974. 



Fork length (cm) of pollock 



pollock. Other fishes identified included gadids, 

 cottids, hexagrammids, and zoarcids. 



Pollock food composition in summer 1974 and 

 spring 1977 can be compared although geographic 

 locations of stomachs collected varied (Figure 4). 

 Pollock were observed with more copepods as a 

 percentage of food biomass in spring 1974 than in 

 summer 1977. Amphipods were nearly absent 

 from stomachs collected in spring 1974 but were 

 an important food component in summer 1977. 



100 



D 



iscussion 



Previous studies on the food of the walleye pol- 

 lock in the eastern Bering Sea indicated that in 

 winter 1972, juvenile pollock fed mainly on 

 euphausiids, while adult pollock fed on 

 euphausiids, small pollock, and other fish (Mito 

 1974). In summer 1970, juvenile pollock fed on 

 copepods and euphausiids, while adult pollock fed 

 on euphausiids, small pollock, and other fish 

 (Takahashi and Yamaguchi 1972). Our study in- 

 dicates that in summer 1974 juveniles fed mostly 

 on copepods, euphausiids, and amphipods, while 

 adults fed on euphausiids, amphipods, and fish. In 

 spring 1977, juvenile pollock fed mostly on 

 copepods and euphausiids, while adult pollock fed 

 on copepods, euphausiids, and fish. The results of 

 these studies indicated that euphausiids are an 

 important year-round food source of both juvenile 

 and adult pollock. Fish appear to be an important 

 year-round resource to adult pollock. The relative 

 importance of other prey organisms in the diet of 

 pollock seems to fluctuate between the studies. 



Adult pollock tend to obtain a greater percen- 

 tage of their food biomass from larger prey or- 

 ganisms than juvenile pollock, by ingesting more 

 fish, euphausiids, and amphipods as they grow 

 larger (Figures 2, 3). Additionally, larger pollock 

 tend to exclude copepods from their diet (Tables 3, 

 4). These observations could result from an active 

 process, based on preference or capture efficiency. 



< 



5 



o 



50 



FISH 



SHRIMP 



DIGESTED MATTER 



ANNELIDS 



CHAETOGNATHS 



EUPHAUSIIDS 



AMPHIPODS 



COPEPODS 



t35cm >35cm 



SUMMER 

 1974 



t35cm >35cm 



SPRING 



1977 



FIGL'RE 4. — Percent biomass of stomach contents by taxa for 

 adult and juvenile walleye pollock in summer 1974 and spring 

 1977 in the Bering Sea. 



or a passive process, resulting from spatial dis- 

 tribution. 



Additional information is needed to understand 

 the complexities of pollock feeding behavior, in- 

 cluding: 1 ) seasonal variations in feeding behavior, 

 2) geographical variations, and 3) effects of alter- 

 nate prey on cannibalism and grazing on other 

 fish. This information would be useful in eco- 

 system modelling to understand the natural com- 

 petitive and predatory interactions between fish 

 populations and the potential effects of heavy 

 exploitation. 



Acknnw Icdgnients 



We thank the following persons at the North- 

 west and Alaska Fisheries Center: Donald Day for 

 collecting and making preliminary analysis of 

 1974 data, Robert French for arranging the collec- 

 tion, and Beverly Vinter and Jay Clark for iden- 



307 



