Prey species - P. caaatsebatica 



o 



5 



1981 



1984 



1985 



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

 Prey species - C. bairdi Prey species - C. opitio 



1984 1985 



1984 1985 



Figure 4. — Percentage of frequency of occurrence (%F0) and percentage by weight (%W) of red king crab, Pamlithodes 

 camtschatica, and two species of snow crab, Chionoecetes bairdi and C. opilio, in stomachs of two Pacific cod size gi'oups in the 

 respective crab consumption area during 1981, 1984, and 1985. 



The only significant relationship for the occur- 

 rence of C. bairdi was fish size; cod 30-59 cm in 

 length contained this crab species more fre- 

 quently than cod >60 cm. The chosen model fits 

 these data poorly because of the presence of one 

 anomalous cell; cod >60 cm in 1981 contained C. 

 bairdi much less frequently than any other 

 group. The anomalous value produces a size-year 

 interaction wherein small cod do not show inter- 

 annual differences in the occurrence of C. bairdi, 

 while large cod seem to have an increase in oc- 

 currence over time. The percentages by weight 

 also show a size-related difference in C bairdi 

 consumption. 



Year was the most important variable describ- 

 ing differences in occurrence of C. opilio in cod 

 stomachs; occurrence increased from 1981 to 

 1985. Size was also significant, with large cod 

 consuming this species more frequently than did 

 small cod. Percentages by weight of C. opilio do 

 not show the same trends as strongly as do the 

 frequency of occurrence data; 1981 and 1984 ap- 

 pear similar and the size-related differences do 

 not look as strong. 



Size Frequencies and Sex Ratios of Crab 

 Species 



Only 10 red king crab CL measurements were 

 taken over all three years because of the ad- 

 vanced state of digestion of most of these crabs 

 found in Pacific cod stomachs. Carapace lengths 

 ranged from 53 to 160 mm with an average of 106 

 mm. Nine out of 10 crabs were larger than 90 



mm. Sex was determined for only one specimen; 

 it was a female. 



The Kolmogorov-Smirnov test (Zar 1974) 

 showed that thei"e was no significant difference 

 iP > 0.05) between C. bairdi size distributions 

 (Fig. 5) consumed by cod 30-59 cm in 1984 and 

 1985, and no significant size difference between 

 C. bairdi consumed by cod >60 cm in 1981, 1984, 

 and 1985. All other size-frequency distribution 

 comparisons, in particular comparisons between 

 cod size groups within years, showed significant 

 differences. Most C. bairdi in smaller cod (30-59 

 cm) were <20 mm CW, while large cod con- 

 sumed crab in the 20-30 mm CW size range. Size 

 distributions of crab <95 mm CW from survey 

 results show size-frequency modes at 40 mm CW 

 or greater. In 1981, most crab in the survey 

 were >60 mm CW. 



The only size-frequency distributions of C. 

 opilio (Fig. 6) that were not significantly differ- 

 ent (P > 0.05) from each other were those for 

 smaller cod (30-59 cm) in 1984 and 1985. The size 

 distributions of C. opilio consumed during 1981 

 by both cod size gi'oups were very different from 

 size distributions for the other 2 years. Smaller 

 cod during that year ate more crabs larger than 

 30 mm CW than in other years, and the modal 

 crab size consumed by large cod (s60) was also 

 greater (40-50 mm CW) compared to 1984 and 

 1985. Crab size distributions from the survey 

 show size frequency modes at 40-50 mm CW. 



The proportions of female snow crabs in Pacific 

 cod stomachs and in resource assessment trawl 

 surveys were determined (Table 3). In cod stom- 



814 



