FISHERY BULLETIN: VOL. 87, NO. 4, 1989 



the size range of red king crab consumed by cod 

 in this study. However, the areas of cod preda- 

 tion on snow crab show interannual variation 

 that does not match areas of adult abundance 

 determined from NMFS trawl surveys, probably 

 because the trawls used in these surveys do not 

 catch small snow crabs <40 mm CW, which is 

 the size consumed most by cod. Surveys showed 

 most adult C. bairdi to be east and southeast of 

 the Pribilof Islands in 1981, 1984, and 1985, 

 while cod stomach contents showed no small C. 

 bairdi east of 165°00'W in 1981. Similarly, sur- 

 veys showed high densities of C. opilio >95 mm 

 CW as far south as the Pribilofs in 1981, whereas 

 most small C. opilio eaten by cod were much 

 farther north in that year. 



The geographic distribution of snow crab pre- 

 dation by Pacific cod was very different in 1981 

 than in the other two years, and the more north- 

 erly location of the ice edge in that year relative 

 to 1984 and 1985 suggests an environmental rela- 

 tionship between C. opilio distribution and 

 physical factors. Somerton (1981, 1982) postu- 

 lated a direct relationship between sjaring ice 

 cover and planktonic larval survival of C. opilio 

 in order to explain observed high recruitment to 

 the adult population of those year classes that 

 may have been in the plankton and benefited 

 from the associated ice edge production during 

 1971 and 1972 in the eastern Bering Sea. Instead 

 of a relationship between ice cover and plank- 

 tonic survival, however, our data suggest that 



ice cover in a given year may also indicate the 

 areal extent of juvenile C. opilio in the same 

 year. Since benthic dwelhng juveniles would not 

 benefit directly from an ice edge bloom, the ice 

 edge in a particular year may be an indicator of 

 another environmental variable such as bottom 

 temperature. 



The average bottom temperature at stations 

 where C. opilio were found in cod stomachs is 

 compared (f-test) with bottom temperatures 

 where no C. opilio were consumed during the 

 three years of this study (Fig. 9). The average 

 bottom temperatures (<3°C) were significantly 

 lower (P < 0.05) for the locations where C. opilio 

 were found than for locations where they were 

 absent. Somerton (1981) reported the weighted 

 average bottom temperatures at stations where 

 C. opilio occurred in 1979 was less than 3°C 

 when the weights used were crab abundance. In 

 that year, juveniles <40 mm CW had the highest 

 abundances per station, indicating that the tem- 

 peratures apply mainly to the juvenile portion of 

 the population. In the northwest Atlantic, 

 Brethes et al. (1987) found bottom tempera- 

 tures <3°C to be the most significant factor in 

 determining the spatial distribution of juvenile 

 C. opilio <40 mm CW in the Gulf of St. Law- 

 rence. Thus, the geogi'aphic distribution of juve- 

 nile C. opilio appears to depend mostly on bot- 

 tom temperature; highest densities are found in 

 areas where bottom temperatures are less than 

 3°C and those areas may be a significant portion 





0) 



O 



Q. 2 



E 

 o 



o 

 m 



C opilio 

 No C opilio 



1984 



Year 



Figure 9. — Mean bottom temperature CO and 95% confidence intervals at 

 locations where Pacific cod consumed Chionoecetes opilio (black circles) and at 

 locations where Pacific cod did not contain C. opilio (open circles) in 1981, 

 1984, and 1985. 



820 



