FISHERY BULLETIN: VOL. 83, NO. 4 



Table 5. — Results of Student-Newman-Keuls multiple range comparisons on head length and 

 gill raker length and spacing of juvenile rockfishes (Sebastes). Numbers in parentheses are 

 significance levels of analyses of variance (ANOVA). 



serranoides, and S. melanops, which all have relative- 

 ly long, slender bodies, had significantly smaller 

 heads than S. caurinus, S. camatus, and S. pinniger 

 (Ikble 5). Sebastes mystinus and S. melanops also had 

 significantly longer gill rakers than S. caurinus and 

 S. camatus, with S. serranoides having intermediate 

 length rakers. Sebastes atrovirens was relatively in- 

 termediate in both measurements, but tended to be 

 more like the predators with larger heads and 

 shorter gill rakers. Groupings for gill raker spacing 

 were less distinct. The open water/substrate- 

 associated classifications also broke down with this 

 measurement. Sebastes pinniger, melanops, and 

 atrovirens had more closely spaced rakers, while S. 

 caurinus and serranoides had the widest spaced 

 rakers. Sebastes mystinus had an intermediate value 

 (Ikble 5). Regression lines obtained from the 

 ANCOVA showed that gape grew isometrically with 

 length in all seven species, but that there were 

 significant differences in the size of gape relative to 

 body size among the species (Ikble 6). The slopes of 

 the lines of In (gape) on In (SL) were all statistically 

 indistinguishable from unity, but the intercepts did 

 show a significant difference (P < 0.001, ANCOVA). 

 Much of this significant difference was likely due to 

 S. caurinus, which had a much larger mouth than 

 the other species. Sebastes serranoides also had a 

 fairly large mouth. All other species had very similar- 

 sized mouths (Ikble 6). 



DISCUSSION 



Many species of fish change diet as they grow (Ross 

 1978; MacPherson 1981). This may be because of 

 changes in habitat, thus changing available food 

 sources, or it may simply be a functional response 



Table 6.— Results of the one way analysis of covariance (ANCOVA) 

 performed on regressions of In (gape) versus In (SL) of juvenile 

 rockfishes. 



to changes in fish size and/or metabolic needs. 

 Although there are dietary differences between the 

 juveniles and adults of the species studied here, their 

 foraging patterns and habitats are very much the 

 same 



Species who were water column feeders as 

 juveniles remain water column feeders as adults. 

 Juvenile S. serranoides ate primarily copepods and 

 zoea larvae (Hobson and Chess 1976; this study). 

 Adults also feed in the water column, eating primar- 

 ily small fish and euphausiids (Love 1978). Juvenile 

 S. melanops fed in the open water (on copepods), as 

 do the adults, which eat small fish, euphausiids, and 

 polychaetes (Hallacher 1978; Roberts 1979). This 

 trend also holds for S. mystinus. Juveniles foraged 

 in the water column, mostly on Oikopleura and 

 copepods. Adults also aggregate and feed in the open 

 water, but eat mostly tunicates and crustaceans 

 (Hallacher 1977; Roberts 1979). 



538 



