BRAY and EBELING: THREE "PICKER-TYPE" FISHES 



Table 7.-Habitat breadths of the three fishes, measured relative 

 to kelp-density, bottom-relief, and position-in-water-column 

 classifications of the cinetransects (see text and Table 2). Sample 

 size is the number of cinetransects from which the species was 

 recorded; S is the number of habitat categories (maximum of 12) 

 in which fish were photographed. See Table 4 for further 

 explanation (note that the nature of habitat breadth precludes 

 sample estimates). 



Species 



Sample size 



e 



B/S 



Kelp perch 

 White seaperch 

 Senorita 



109 



42 



201 



11 

 12 

 12 



6.32 

 7.61 

 9.84 



0.57 

 0.63 

 0.82 



most in total resource use, with food the main 

 contributor (Table 8). Yet their large food overlap 

 was caused not by any overall similarity in dietary 

 arrays, but by their sharing one predominating 

 item, the plant-encrusting bryozoans. In fact, rank 

 orders of their food items were not significantly 

 correlated (tau = 0.20, P = 0.16). 



The kelp perch had the least amount of total 

 overlap with others. Sharing the kelp-canopy area 

 to a great extent, the kelp perch and senorita 

 overlapped most in habitat even though the 

 senorita had the broader overall spatial distribu- 

 tion within the kelp bed. Also, the two species' 

 small food overlap and different activity patterns 

 tended to minimize their total overlap. Actually, 

 rank orders of their food items were significantly 

 correlated (tau = 0.50, P<0.001), because the two 

 species shared similar proportions of a number of 

 minor items. Yet overlap was small because they 

 did not share the same predominating item. A low 

 amount of overlap in total resource use was shown 

 by the two perches. Although their activity pat- 

 terns were similar, their diets and habitat 

 preferences differed markedly. In diet, they 

 shared neither a predominating food item nor an 

 array of minor items, and rank orders of their 

 items were uncorrelated (tau = 0.06). 



DISCUSSION 



The kelp perch and senorita-the two species 



most often involved in cleaning activity— co-occur 

 to a great extent in the sunlit upper waters and eat 

 few if any benthic prey. Kelp perch typically feed 

 in loose aggregations of a few to over 30 in- 

 dividuals. Constantly changing direction and 

 depth, feeding individuals flit about to pick par- 

 ticles from mid-water or, occasionally, from the 

 surfaces of kelp and from other fishes. In calm, 

 clear water, these aggregations often extend to 

 the more open areas between kelp plants. In 

 strong currents, however, the fishes gather in back 

 of kelp columns where the water is quieter and 

 where food swept off the surfaces of kelp may be 

 consumed. Solitary senoritas occasionally nip at 

 various substrates and large drifting particles, but 

 they feed most intensely when in large schools. 

 These schools move in and about kelp stands, 

 momentarily dispersing for individuals to pick and 

 tear at kelp fronds and encrustations, then re-as- 

 sembling and moving on to another stand of kelp. 

 The habit of kelp perch and senoritas of cleaning 

 other mid-water fishes probably is incidental to 

 the co-occurrence of the two species in the kelp 

 canopy. Cleaning is not the principal occupation of 

 either species (Limbaugh 1961, Hobson 1971). 

 Their presence in the canopy better relates to their 

 ability to pick small prey from off and from about 

 the kelp blades. Seldom straying far from the 

 heavy foliage where prey may become densely 

 concentrated (Wing and Clendenning 1971), kelp 

 perch also select plankton from incoming currents. 

 Like other diurnal planktivores (Hobson 1974), the 

 kelp perch has a relatively slender body, deeply 

 forked caudal fin, and a slightly upturned mouth 

 (Hobson 1971). Senoritas, which range more 

 widely in the water column from canopy to near 

 bottom, eat much less plankton. They favor at- 

 tached food, primarily plant-encrusting 

 bryozoans, either from the drift or torn from liv- 

 ing plants. 



White seaperch, which usually range nearer the 

 bottom and only clean occasionally (Hobson 1971), 



Table 8.— Overlap in resource use between members of all pairs of the three fishes. Activity overlap is the mean of two independent 

 estimates: from feeding bites observed in the field, and from swimming movements observed in the laboratory. Habitat overlap is 

 measured relative to kelp-density, bottom-relief, and depth classifications of the cinetransects (see Table 3). Total overlap is 

 somewhere between the minimum and maximum estimates. 



825 



