BRAY: INFLUENCE OF WATER CURRENTS AND ZOOPLANKTON DENSITIES 



fish, 0.70, was significantly greater than that for 

 the 14 excurrent fish, 0.20 (Mann-Whitney (/-test, 

 P<0.001). 



Nine categories of food items were identified 

 from gut contents of caged blacksmith (Table 5). 

 Larvaceans and copepods predominated, while 

 other groups were usually rare or absent. All 

 items were typically planktonic except for the 

 sessile stage of the hydrozoan, Obelia sp., which 

 quickly colonized cages even though they were 

 scrubbed before each experiment. Obelia sp. oc- 

 curred in gut contents as small branches (<5 mm 

 long), and each was counted as one individual. 

 Eliminating Obelia sp. would decrease gut full- 

 ness for excurrent fish even more (Table 5). 

 Dietary variation between cages included differ- 

 ences in relative abundances of larvaceans and 

 copepods. and additions of rare items in the 

 excurrent cages. In the incurrent cage, larvaceans 

 were the most abundant food items in all experi- 

 ments, but in the excurrent cage, they were most 

 abundant in only two experiments. When present, 

 large copepods (>4 mm) were found mostly in 

 incurrent-caged fish. Though relatively few in 

 numbers, the size of these copepods (some nearly 

 10 mm long) probably made them nutritionally 

 important. That excurrent fish ate Obelia sp. 

 during three experiments is difficult to explain. 

 Tufts of Obelia sp. may have been more abundant 

 in the excurrent cages, because the excurrent 

 cages appeared to foul at a faster rate. Sessile 

 hydroids are not a normal food of blacksmith 

 (Hobson and Chess 1976). 



Caging altered the blacksmith diets, but the 

 effects were variable. In two of three collections, 

 free fish had generally consumed more food than 

 caged fish (Table 6). Seven categories of food items 

 were identified in the guts of caged and free fish. 

 As before, larvaceans and copepods were by far the 

 most abundant. Cladocerans were abundant in a 

 few individuals, but chaetognaths, decapod larvae, 

 and polychaetes were uncommon. Free fish ate 

 mostly larvaceans in all of the collections, but 

 caged fish were inconsistent. In one collection, 

 caged fish ate mostly large copepods, but in the 

 other two, they ate mostly larvaceans. 



Free-living juveniles at the incurrent end ate 

 more food than those at the excurrent end (Table 

 7). Pooled among collections, gut fullness differed 

 significantly between reef ends (Mann-Whitney 

 (/-test, P<0.05). Nonetheless, dietary compo- 

 sition of all free-living juveniles was similar. 

 Larvaceans always made up the most abundant 



item, with copepods and cladocerans also common. 

 Numbers of small copepods were slightly less in 

 excurrent fish, but the difference was not nearly so 

 great as for larvaceans or large copepods. Num- 

 bers of cladocerans were greater in excurrent fish. 



DISCUSSION 



Blacksmith Distribution Patterns 

 Adults 



The midwater surveys indicate that large num- 

 bers of adult blacksmith (>150 mm TLi swim 

 to the incurrent end of Naples Reef. Under the 

 usual current pattern of flow from the east, 

 almost all adults recorded were at the east end; 

 when currents reversed, adults were far more 

 abundant at the west end. During one survey, 

 adults were actually seen migrating to the oppo- 

 site end as currents reversed. The only times I saw 

 large numbers of adults dispersed throughout the 

 reef occurred when currents were negligible. On 

 another occasion at To yon Bay, Santa Catalina 

 Island (190 km southwest of Naples Reef), I saw 

 a similar response of blacksmith to a current 

 reversal. 



Observations at night indicate that large num- 

 bers of blacksmith of all sizes take shelter in holes 

 at the west (usually the excurrent) end of Naples 

 Reef. Indeed, the density of sheltering blacksmith 

 at the west end may exceed that at the east 

 because higher rocky relief and more complex 

 substratum at the west end provide more refuges. 

 An investigation of the sheltering behavior of 

 tagged blacksmith indicated that many individ- 

 uals tend to return to the same shelter at night 

 (Bray in prep.). Yet when the current flowed from 

 the east, extensive searches throughout the entire 

 reef during the day failed to reveal substantial 

 numbers of adults anywhere but at the east end. 

 During the present midwater surveys, I saw one of 

 these tagged fish in a feeding aggregation at the 

 extreme eastern margin of the kelp, almost 300 m 

 away from the hole where it was tagged. This, and 

 my observation that blacksmith swam the length 

 of the reef when currents reversed, indicates that 

 some adults must swim a considerable distance 

 each day to gather at the incurrent end. 



Juveniles 



In contrast, juvenile blacksmith (<125 mm TL) 



837 



