HOBSON: CLEANING SYMBIOSIS 



seiiorita and drive it away. The rubberlip perch 

 was noted taking the latter course of action per- 

 haps more often than the other species, even 

 considering the relatively few times it was ob- 

 served being cleaned. 



Although the kelpfish is the only species that 

 was observed consistently initiating cleaning 

 from seiioritas, individuals of several species do 

 so in at least one situation. This occurs where 

 exceptionally large concentrations of senoritas, 

 sometimes thousands, swim above the rocks. 

 At various times, garibaldis, pile perch, rubbei'- 

 lip perch, olive rockfish, and others were ob- 

 served hovering in the soliciting posture amid 

 these concentrations (Figure (5) until one of the 

 senoritas approached and cleaned them. 



Material Removed from Other Fishes by 

 the Senorita 



Food habits of cleaning senoritas. — An obvi- 

 ous question is: What do seiioritas remove from 

 the bodies of fishes they clean? Limbaugh 

 (19.55) stated that senoritas remove bacteria. 



parasitic copepods, and isopods. He was not 

 more si)ecific than this, nor did he present data. 

 Various other cleaners reportedly take not only 

 ectoparasites but also diseased and neci'otic tis- 

 sue (Feder, 1966, and others.) 



To determine just what it is that sefioritas 

 remove from the bodies of other fishes, I ex- 

 amined the gut contents of 27 specimens. 111 to 

 175 mm long, that were speared while they were 

 cleaning other fishes. Food items in their guts, 

 ranked as percentage of each item in the entire 

 sample, were as follows: caligid copepods, 39% ; 

 gnathiid isopod larvae, 12 S' ; algae with en- 

 crusting bryozoans, 10 /r ; caprellid amphipods, 

 5',t ; fish scales, iVr; and fragments of nonpar- 

 asitic crustaceans, 4',i . Unidentified material 

 made up 26 "^r of the sample. Of the 27 speci- 

 mens, ectoparasites occurred among the gut con- 

 tents of all but two. In most, the ectoparasites 

 predominated. Even though the data are con- 

 vincing, they do not fully reflect the extent to 

 which cleaning oliviously dominated the activity 

 of these particular fish for at least several hours 

 leading u]) to their capture. This is because the 



Figure 6. — Garibaldi hovering amid a large assemblage of senoritas. 



503 



