404 



Fishery Bulletin 102(2) 



shallower areas in the Pavlof Bay area <Fig. 6). From the 

 shrimp survey data, Anderson (2000) showed that pan- 

 dalid shrimp occupying inshore and shallower water (e.g., 

 Pavlof Bay area) declined to near extinction (<0.1 kg/km) 

 from 1978 to 1982, while offshore and deepwater pandalid 



DAug-80 



 May-81 

 QSep-81 



 Aug-95 



Main prey items 



Figure 4 



Percentage by volume (for the values in 1980s) and the percentage by weight 

 (for the values in 1995) of the main prey items of Pacific cod collected in Pavlof 

 Bay. Alaska. 



a 



Main prey items 



Figure 5 



Percentage by weight of the main prey items in the diet of Pacific cod collected 

 in 1990 (GOA90), 1993 (GOA93), and 1996 (GOA96) in the Gulf of Alaska and 

 in 1995 (PAV95) in Pavlof Bav. 



shrimp species maintained low population levels (>0.1 



kg/km ). The data from this study corroborates Anderson's 



(2000) results. 

 Anderson (2000) also reported that during the period 



of the decline of pandalid shrimp in inshore waters of the 

 Gulf of Alaska, the abundance of some 

 pleuronectids, Pacific cod, and walleye 

 pollock increased. These species are 

 predators of pandalid shrimp (Yang 

 and Nelson, 2000). One hypothesis is 

 that predators keep pandalid shrimp 

 populations low. Albers and Anderson 

 (1985) suggested that cod predation 

 was one reason for the failure of the 

 pink shrimp stock to rebuild in Pavlof 

 Bay. In the Northwest Atlantic. Lilly 

 et al. (2000) showed that the large in- 

 crease in shrimp biomass seen in the 

 1990s was related to the collapse of cod 

 \Gctdus morhua) populations during the 

 late 1980s and 1990s in the northeast 

 Newfoundland shelf. The impact of cod 

 on Barents Sea shrimp (P. borealis) 

 was also reported by Berenboim et 

 al. (2000). They found that when cod 

 biomass is high, the shrimp frequency 

 of occurrence in cod stomachs declines; 

 there is a significant inverse correla- 

 tion between the abundance of cod and 

 shrimp. 



Tanner crabs consumed by Pacific cod 

 in this study ranged from 5 to 42 mm 

 carapace width (CW). In general, the 

 size of Tanner crabs consumed in- 

 creases as Pacific cod size increases. 

 The size range of Tanner crabs con- 

 sumed by Pacific cod in this study is 

 similar to that (5-45 mm) found in 

 Pacific cod stomachs in Albers and An- 

 derson's ( 1985 ) study and is also similar 

 to that (1-40 mm) found in Hunter's 

 (1979) study near Kodiak Island. 



Jewett's ( 1978 ) Pacific cod diet study 

 around Kodiak Island from 1973 to 

 1976 showed that Tanner crabs were 

 the most frequent (37%) prey of Pa- 

 cific cod; pandalid shrimp occurred in 

 8-10% of the stomachs examined from 

 1973 to 1975; and walleye pollock were 

 found in 49r of the stomachs examined. 

 The importance of Tanner crabs as food 

 of Pacific cod in Jewett's ( 1978 1 study is 

 coincident with our study. 



This study suggests that there were 

 substantial differences between the 

 diets of Pacific cod in Pavlof Bay be- 

 tween the early 1980s and 1995. In 

 the 1980s, pandalid shrimp and cap- 

 elin were the main food of Pacific cod, 

 whereas benthic species (polychaetes, 



