SINGER: FOOD HABITS OF ROCKFISH 



The ontogenetic similarity in foraging was also 

 seen in substrate-oriented feeders. Juvenile S. 

 atrovirens ate mostly large, demersal, gammarid am- 

 phipods. However, copepods ranked second in their 

 diet, presumably because of small individuals feeding 

 in the open water below the kelp canopy. Adults of 

 this species feed on amphipods, shrimp, and crabs 

 which can be either demersal or free swimming 

 (Quast 1968; Hobson and Chess 1976; Roberts 1979). 

 Small juvenile S. caurinus and 5. carnatus fed in 

 the open water, while larger juveniles foraged demer- 

 sally. The adults of these species also feed demer- 

 sally. Sebastes carnatus is known to eat juvenile 

 rockfish, ophiuroids, and crustaceans (Hallacher 

 1977; Roberts 1979), while adult S. caurirms eat 

 mostly brachyurans and shrimp (Prince 1975; Prince 

 and Gotshall 1976). Juveniles also seem to exhibit 

 the same microhabitat preference as adults. Sebastes 

 serranoides, mystinus, and melanops were found 

 mainly in the midwater, while S. caurinus, carnatus, 

 and atrovirens were generally seen to associate more 

 closely with some physical substrate (kelp plants, 

 rocks, etc.) (Carr 1983; pers. obs.). Thus, in these 

 species it seems that once an individual has survived 

 its life as a pelagic larva and entered the kelp forest, 

 it assumes the general habitat and foraging 

 characteristics of an adult. 



Sebastes pinniger was the only species which show- 

 ed different foraging patterns between juveniles and 

 adults. Juvenile 5. pinniger were generally found 

 close to the bottom over sand or in association with 

 the rock/sand interface at the edge of the reef (Carr 

 1983; pers. obs.) while adults occur higher in the 

 water column in deep water offshor& Juveniles fed 

 demersally on copepods over sand and drift algae 

 very near the kelp forest. Adults feed in the water 

 column on euphausiids and small fish (Phillips 1964). 

 Recently, adults have been found to be more demer- 

 sal feeders than other offshore rockfish (Brodeur 

 1982). 



In assessing the mechanisms which might lead to 

 the observed diet differences, several factors must 

 be considered, such as prey distribution and abun- 

 dance, prey availability, predator morphology, prey 

 and predator activity patterns, and predator 

 distribution. 



Qualitative analysis of plankton samples, combined 

 with underwater observations, showed that plankton 

 were very abundant. Although the diel behavior pat- 

 terns of shallow, inshore zooplankton are highly 

 variable with respect to specific habitats, species, 

 seasons, and latitudes, the patterns of plankton 

 distribution observed were quite comparable with 

 reported accounts in kelp forests (Hobson and Chess 



1976; Coyer 1979; Hammer 1981). 



Predator morphology is an important factor deter- 

 mining prey size in planktivorous fish. Certain 

 features distinguish water column foraging fish, such 

 as long, slender bodies, sharp head profiles, fine den- 

 tition on jaws and pharyngeal bones, and long, close- 

 ly spaced gill rakers (Yasuda 1960; Davis and Bird- 

 song 1973). Of the seven species studied, the water 

 column feeders— 5. mystinus, S. serranoides, and S. 

 melanops— ha.d smaller heads with longer gill rakers. 



Water column foragers and substrate-associated 

 foragers could be separated by both morphology and 

 prey siza The water column foragers— S. mystinus, 

 S. serranoides, and S. melanops— a.\\ had relatively 

 long, slender bodies, small heads, and long gill 

 rakers. This agrees with predictions, especially since 

 the adults of these species are also water column 

 foragers. Water column feeders also ate substantially 

 smaller prey than did substrate predators (see Fig- 

 ure 2, Tkble 2). The substrate-associated feeders— 

 S. caurinus, S. carnatus, and 5. atrovirens, which 

 ate larger prey— all had stouter bodies, larger heads, 

 and shorter, less ornamented gill rakers. 



Juveniles of S. pinniger were somewhat inter- 

 mediate in feeding morphology. Their heads were 

 large and stout, indicative of substrate-oriented 

 feeding. However, they had long, thin bodies and 

 long, fairly closely spaced gill rakers, which is in- 

 dicative of water column foraging, as is displayed by 

 adults. This intermediate situation may be indicative 

 of the fact that S. pinniger may go through a second 

 ecological transition from a reef-dwelling, substrate- 

 associated juvenile form back to a more pelagic off- 

 shore situation as an adult. 



Differing diel patterns can be one way for co- 

 occurring predators to exploit similar resources 

 while keeping interspecific interactions low (Keast 

 and Webb 1966; Schoener 1974; Bray and Ebeling 

 1975). In zooplanktivores diel patterns of both 

 predators and prey are important (Hobson and Chess 

 1976; Robertson and Howard 1978). Most juveniles 

 were active only during the day, with the possible 

 exceptions of S. pinniger and S. serranoides (Singer 

 1982). This was also reflected in greater stomach 

 fullness of most species in the afternoon and early 

 evening, indicating diurnal feeding patterns (Singer 

 1982). Plankton abundances were high during both 

 day and night. However, more species were found 

 in the water column at night. Thus, while some of 

 these fish could do well feeding at night, daytime 

 abundances of food seemed sufficient for their needs. 

 Juveniles were indeed found to be most active and 

 to feed most frequently during daylight hours, but 

 intraspecific variability was high (Singer 1982). 



539 



