FISHERY BULLETIN: VOL. 82, NO. 2 



TABLE 6. — Summary of Pacific ocean perch, Sebastes alutus, 

 stomach contents from the National Marine Fisheries Service's 

 summer sampling. F.O. = frequency of occurrence. 



TABLE 7. — Summary of splitnose rockfish, Sebastes diploproa, 

 stomach contents from the National Marine Fisheries Service's 

 summer sampling. F.O. = frequency of occurrence. 



Number of 

 prey 



Weight of 

 prey (g) 



(B), and had the most even distribution among 

 prey types (B n ) of all rockfish examined from the 

 summer survey. Sebastes diploproa preyed on 

 fewer taxa than S. flavidus but had moderately 

 high overall and normalized food breadth values. 

 Sebastes pinniger, S. crameri, and S. alutus uti- 

 lized a similar number of distinct prey items and 

 had similar breadth and evenness values with S. 

 alutus having a more equitable distribution of 

 prey than the other two. 



The seasonal results for the S. flavidus and S. 

 pinniger were more divergent and represent the 

 extreme values found among the species. Seven- 

 teen principal prey types were important in the 

 seasonal diet of S. flavidus, contributing toward 

 a high B value. However, the dominance of a 

 few species yielded a low evenness value for this 

 species. Sebastes pinniger preyed on few taxa 

 in fairly unequal proportions yielding fairly low 

 niche breadth and evenness values. These low 

 evenness values could be caused by the prepon- 

 derance of euphausiids found in the guts of both 

 species during the summer months. 



The individual overlap coefficients and the 

 mean overlap for each species are presented for 



Number of 

 prey 



Weight of 

 prey (g) 



both the weight and numerical abundance of prey 

 in Table 10 for the summer surveys. As overlap 

 indices are affected by the level of taxonomic 

 specificity at which the prey have been identified, 

 no unbiased means for testing the significance of 

 these values are available. We adopted the con- 

 vention that overlap values from 0.00 to 0.29 are 

 considered low, 0.30 to 0.60 considered medium, 

 and those above 0.60 show highly similar diets 

 (Langton 1982). 



The coefficients for numerical composition show 

 high values for all possible combinations except 

 those involving S. crameri. Very similar propor- 

 tions of the major euphausiid prey groups resulted 

 in an extremely high overlap value (0.93) between 



278 



