BRODEUR and PEARCY: FOOD HABITS AND DIETARY OVERLAP OF SOME SEBASTES 



ratios at the individual times were subjected to an 

 analysis of covariance which compared the weight 

 ratios adjusted for fish size (Jenkins and Green 

 1977). Both S. pinniger (F (12 , 3 54> : 5.68, P < 

 0.001) and S. flavidus (F( U ,262) = 6.51, P < 0.001) 

 showed significant differences in the mean weight 

 ratios over the times tested, implying that feeding 

 varied during the diel period. No significant dif- 

 ferences (P > 0.05) in stomach fullness were 

 associated with size or sex of the predator for 

 either species. 



Predator-Size Variation 



The proportion of empty S. pinniger stomachs 

 as well as the percent frequency of occurrence and 

 percent weight of prey taxa were remarkably 

 invariant among the four predator size classes 

 (Table 14). Only the largest size class (2:55+ cm) 

 shows any substantial variation with a larger pro- 

 portion by weight of fishes and a commensurate 

 decrease in weight of euphausiids consumed. 

 Much of this fish weight was contributed by a few 

 individual fish of large relative size (mostly adult 

 S. jordam); the frequency of occurrence of fishes 

 is only slightly higher for this largest size class. 



Few obvious size-related trends were apparent 

 for S. flavidus. The two smallest size classes 

 consumed the largest proportion of euphausiids. 

 Euphausia pacifica were less important for large 

 fish. Decapods and cephalopods showed similar 

 trends except that the frequencies of occurrence 

 were highest for cephalopods but lowest for deca- 

 pods in the largest size class. Fishes were consis- 

 tent in their weight and occurrence proportions 



except that one size class (40-<45 cm) had much 

 lower proportions than the others. Few trends 

 were apparent for either amphipods or gelatinous 

 zooplankton although both groups were commonly 

 found. 



To determine if different sizes of rockfish se- 

 lected different sizes of prey, all fish that con- 

 tained measurable prey were grouped into 10 mm 

 length intervals and the means and ranges of their 

 prey were plotted against fish size (Fig. 8). Al- 

 though some exceptions exist, the majority of the 

 prey of S. pinniger are found within a narrow 

 range of prey sizes, a range (15-27 mm) largely 

 determined by adult euphausiids, the dominant 

 prey category (Fig. 8). Fishes of the largest two 

 size classes consumed larger prey on average, and 

 their prey had the largest variation in size due to 

 high numbers of both small and large prey con- 

 sumed by these fish. No significant relationship 

 was found between length of fish and either 

 overall size of prey or size of euphausiid prey. 



Sebastes flavidus showed a much greater range 

 in the sizes of prey consumed with the variation 

 and range in prey length increasing with size 

 of predator (Fig. 8). The mean size of prey eaten 

 did not appreciably increase until the very largest 

 size classes. Although the maximum prey size 

 increases with fish size, the minimum size varies 

 little throughout the length ranges examined. 

 Again for this species, no relationship was found 

 between fish length and overall or euphausiid 

 prey lengths. 



The size distribution of prey is shown for both 

 species in Figure 9. The prey-size spectrum of S. 

 pinniger was distributed fairly normally with the 



TABLE 14. — Variation in major prey taxa composition with size of predator for Sebastes pinniger and S. flavidus. F.O. = frequency 

 of occurrence; '/ W = percent gravimetric composition; + = a prey category was present but made up < O.l'X of the total weight. 



285 



