FISHERY BULLETIN: VOL. 72, NO. 4 



monacanthids and tetraodontids (all Tetraodon- 

 tiformes). In pointing out that coelenterates are 

 not food for fishes in most marine communities, 

 Hiatt and Strasburg (1960) cited various 

 specialized features of fishes that prey on corals in 

 the Marshall Islands: for chaetodontids and 

 monacanthids that snip off individual polj^DS, they 

 listed the produced snouts, small terminal 

 mouths, and fine protruding incisiform teeth; for 

 tetraodontids and balistids that bite off larger 

 pieces of coral, they noted very heavy, strong den- 

 tition. All fishes that feed on coral, including 

 those that feed heavily on coral mucus, seem to be 

 diurnal. Obviously a predator that bites off large 

 chunks of coral, or w^hich scrapes away mucus, 

 would find diurnal habits adaptive — its food is 

 equally accessible day or night, and its own activ- 

 ity would benefit from daylight. On the other 

 hand, the polyps of some coral species are most 

 expanded at night, suggesting that perhaps pred- 

 ators that would snip them off might find them 

 most accessible after dark; however, the precise 

 manipulations involved in this activity probably 

 require the light of day , because without exception 

 all such predators are diurnal. 



Daylight and precise manipulations also seem 

 required of predators that pluck tiny cryptic or- 

 ganisms, notably amphipods, from amid benthic 

 cover. An example from Kona is the labrid 

 Anampses cuvier, whose prey are amphipods and 

 other organisms too small for large-mouthed 

 generalized predators of comparable size. Taking 

 such prey requires a specialized tactic and feeding 

 mechanism. Characteristically such predators 

 hover within a few centimeters of the substratum, 

 inspecting the surface. When they spot prey 

 — perhaps through movement or an unusual 

 contour — they take it in a characteristic plucking 

 manner. 



Probably this way of plucking tiny prey from a 

 substratum preadapted precursors of those fishes 

 that are specialized as cleaners. Most cleaner 

 fishes, which include certain labrids, pomacen- 

 trids, and chaetodontids, pluck various materials, 

 mostly ectoparasitic crustaceans, from the bodies 

 of other fishes. Possessing both the necessary 

 techniques and morphology, certain fishes in this 

 category were prepared to adopt the cleaning 

 habit when their concept of a suitable feeding sub- 

 stratum broadened to include the bodies of other 

 fishes (Hobson, 1971). A few species, like 

 Labroides phthirophagus in Kona, are specialized 

 as cleaners, having refined both their feeding 



morphologies and techniques to more efficiently 

 practice this habit. All known cleaner fishes are 

 diurnal. 



Most of the invertebrate prey of diurnal fishes 

 are insignificant as prey of nocturnal fishes. How- 

 ever, the specializations that permit certain diur- 

 nal fishes to seek out secretive prey in daylight 

 make available to them at that time some of the 

 forms — motile crustaceans in particular — that 

 are important prey of various generalized pred- 

 ators after dark. For some fishes, the adaptations 

 that permit them to take crustaceans and other 

 forms from under reef cover in daylight are mor- 

 phological. Thus, the chaetodontid Forcipiger 

 longirostris and the labrid Gomphosus varius both 

 have elongated snouts with which they reach deep 

 into reef crevices for crustaceans. In other fishes 

 the adaptations that make secretive prey avail- 

 able are more strictly behavioral. Thus, the labrid 

 Thalassoma duperrey follows close to the feeding 

 jaws of large herbivores and other fishes that dis- 

 turb the substratum, and snaps up tiny crusta- 

 ceans driven from cover. This behavior is wide- 

 spread, occurring in other wrasses in Kona and 

 also in the Gulf of California (Hobson, 1968a). 

 Some species lower on the evolutionary scale seem 

 to have similar behavior: as suggested above, the 

 carangid Caranx melampygus may enjoy this ad- 

 vantage by following the mullid Parupeneus 

 chryseydros, as may the aulostomid Aulostomus 

 chinensis by accompanying grazing schools of 

 acanthurids — in these two situations, however, 

 the prey seem to be mostly small fishes. 



Some diurnal predators excavate buried prey, 

 as when the labrid Coris gainiard overturns small 

 stones with its snout and feeds on animals thus 

 exposed. And in the eastern Pacific the balistid 

 Sufflamen verres uncovers prey buried in the sand 

 by exposing them with a jet of water from its 

 mouth, or by rapidly undulating dorsal and anal 

 fins while lying on its side, thereby generating 

 currents that sweep the sand away (Hobson, 1965, 

 1968a). Similarly, the ostraciontid Lactophrys 

 triqueter in the tropical Atlantic by jetting water 

 from its mouth uncovers prey buried in the sand 

 (Longley, 1927). 



Related Problems of Species Recognition. — The 

 enormous potential for varied feeding adaptations 

 in these advanced teleostean groups has led to the 

 occurrence on most coral reefs of large numbers of 

 closely related species that seem to have diverged 

 from one another chiefly on the basis of differing 

 food habits. For example, 14 species of the genus 



1026 



