FISHERY BULLETIN: VOL. 82, NO. 2 



150-7 

 125- 



E 



E 100 H 



UJ 

 M 



OT 75H 



>- 



Q- 504 



25- 



0- 



400 



S. pinniger 



U?t 



)ii> 



f^ + i |t 



llll 



450 



500 



—I — i — i — •- 



550 



600 



FIGURE 8. — Mean (horizontal lines) ±95% confidence 

 limits (boxes) and ranges (vertical lines) of prey sizes 

 found for each 10 mm interval of Sebastes pinniger and 

 S. flavidus. 



1507 



125- 



100- 



<" 75 

 >- 



UJ 



cr 



Q- 50 



25- 



S flavidus 



it^nHt 





iiim 



«fi 



.mi 



} i i i I I « i I I i ' I i  I I I ' I I I I I I I i I t I I 



300 350 400 450 500 550 600 



PREDATOR LENGTH (mm) 



mode coinciding with the mean (x = 10.38 mm), 

 although disjunct groups of small and large prey 

 were found (Fig. 9). The prey-size spectrum of S. 

 flavidus was slightly skewed toward the larger 

 sized prey with the mean size (x = 18.44 mm) less 

 than the mode. A smaller peak also appeared 

 around 25 mm. No significant differences were 

 found in the mean prey sizes utilized by the two 

 species (Student's f-test, P > 0.05). 



Analysis of Variation 



The results of the chi-square analyses for S. 

 pinniger showed that none of the factors analyzed 

 had a significant effect on the occurrence of food in 

 the stomachs (Table 15). At least one of the factors 

 was related to the occurrences of all seven prey 

 categories examined. Seasonal effects were the 

 most significant (all P ^ 0.01) and were due to the 

 higher occurrences of hyperiid amphipods, fishes, 

 and gelatinous zooplankton in fall and winter. 

 Area and time of capture showed both highly 

 significant (P s 0.001) and insignificant effects 



depending on the prey category, but most compar- 

 isons were significant at the 0.05 level. In none of 

 the prey categories examined did the size of the 

 predator have a significant effect on the relative 

 proportions consumed. 



For S. flavidus, season of capture and size of 

 predator affected the proportion of empty stom- 

 achs found (Table 15). Again season had the most 

 significant influence on prey occurrence and was 

 significant in all eight prey categories. Highly 

 significant differences were found in area of cap- 

 ture and size of predator especially in the euphau- 

 siid and fish categories. Differences in occurrence 

 of prey with time of capture deviated from ex- 

 pected the least of all the factors analyzed. 



DISCUSSION 



The five species of rockfishes examined rely 

 heavily, if not exclusively, on pelagic macrozoo- 

 plankton and micronekton. Although some ben- 

 thic species appear in the prey lists (e.g., Lyopsetta 

 exilis, Munida quadrispina, Psettichthys melan- 



286 



