FISHERY BULLETIN: VOL. 78, NO. 4 



apparently do not congregate at the incurrent 

 end. Although midwater counts were highest at 

 the excurrent end, bottom observations indicate 

 that juveniles occur abundantly throughout the 

 reef. Some form large stationary aggregations 

 about reef prominences while others are more 

 dispersed, hovering within a few meters of the 

 rocky substratum. Halfgrown fish are most abun- 

 dant along the reef edge, between aggregations of 

 adults and juveniles. To simplify the transects, 

 I tallied blacksmith as though they were com- 

 prised of two major size classes, juveniles and 

 adults; halfgrown fish made up but a small group 

 of intermediate-sized individuals that allowed 

 clearer distinction between these two classes. 

 Actually, fish sizes ranged almost continuously 

 from small juveniles to large adults. The degree of 

 fish movements vary accordingly, from very short 

 forays of newly settled juveniles to extensive 

 migrations of large adults. 



Foraging at Incurrent End of the Reef 



Adults 



In synthesizing day and night observations of 

 fish residing on temperate and tropical Pacific 

 reefs, Hobson (1973) concluded that when fish are 

 active their dominant behavior is feeding, and 

 when they are inactive they seek security either 

 by schooling or by sheltering. Hobson (1972) 

 states, "A suitable feeding location for any given 

 species may or may not be near areas that offer it 

 suitable security during its inactive period. Con- 

 sequently, the major actions of these fishes char- 

 acteristic of twilight relate to moving between 

 feeding locations and shelter locations." 



The most suitable foraging site for adult black- 

 smith, in terms of food availability, is likely at the 

 incurrent end of the reef. The paired caging 

 experiments indicated that the amount of zoo- 

 plankton consumed by adults at the incurrent end 

 was greater than the amount eaten by those over 

 the reef near the excurrent end. Although the 

 caging procedure itself did influence blacksmith 

 foraging, I assume the effect was similar in both 

 cages, so the differences in gut fullness reflected 

 the relative availability of food at the reef ends. 



There are at least two possible reasons for the 

 greater food abundance at the incurrent end. 

 First, plankton is probably replenished there at a 

 faster rate. Measurements of current velocities 



indicated that water crossing the reef is slowed 

 and deflected by rocky prominences and columns 

 of giant kelp. When feeding in a current, black- 

 smith often position themselves in areas of slack 

 water behind kelp while currents deliver food 

 (Hobson and Chess 1976). On the other hand, fish 

 in relatively calm water at the excurrent end may 

 have to swim about, possibly farther from kelp or 

 other shelter, to encounter food at a comparable 

 rate. Hobson and Chess (1976) observed that 

 feeding rates of blacksmith were greater in a 

 moderate than a slack current. 



Second, the density of zooplankton is greater at 

 the incurrent end. Under the normal pattern of 

 current flow with water coming from the east, 

 zooplankton densities were consistently greater at 

 the east end of the reef. Even more convincing, 

 however, was the effect of current reversal; on 

 these two occasions, zooplankton densities were 

 significantly greater at the west end, with most of 

 the differences attributable to decreased densities 

 of cladocerans, small copepods, and larvaceans. I 

 feel that the plankton samples provided a good 

 measure of the abundance of the blacksmith's 

 potential prey because the most abundant items in 

 the plankton samples (copepods, cladocerans, and 

 larvaceans) were also the major items found in the 

 blacksmith guts. Also, I sampled in areas where 

 blacksmith normally gather in the appropriate 

 current conditions, and I was invariably sur- 

 rounded by foraging adults while I collected 

 plankton at the incurrent end. Although several 

 investigators have discussed decreased densities 

 of plankton in kelp beds (Limbaugh 1955; Quast 

 1968b; Miller and Geibel 1973; Feder et al. 1974), 

 to my knowledge, this is the first quantitative 

 documentation. 



Juveniles 



The incurrent end of the reef would seem to be 

 the most suitable foraging site for juveniles — at 

 least those that forage here consume more prey, as 

 determined by the caging experiments and exam- 

 inations of free-living individuals. But the sur- 

 veys showed that juveniles fail to concentrate 

 here, which indicates that other factors override 

 the advantages of incurrent foraging. 



Optimization models may be used to inter- 

 pret foraging movements of planktivorous fishes 

 (Reese 1978). While the benefits of movements 

 usually involve energy gains, costs may include a 

 variety of factors, such as competition, expendi- 



838 



