FISHERY BULLETIN: VOL. 70, NO. 3 



is detailed elsewhere (Hobson, in preparation). 

 The occasional occurrence during twilight of 

 large barracuda. Sphyraena barracuda, and 

 sharks, especially Carch^xrhinus milberti, would 

 not seem to have much impact on the general 

 behavior patterns of the reef fishes. However, 

 any behavior that reduces vulnerability to pred- 

 ators certainly would enhance survival on the 

 occasions when such predators do appear, and 

 the effect of this circumstance may be greater 

 than is readily apparent. Nevertheless, with 

 neither schooling prey nor the predators that 

 exploit them being major elements of the fauna, 

 the vigorous interactions such fishes so promi- 

 nently generate during twilight in the Gulf of 

 California are essentially missing in Kona. 



The absence of large piscivores on the Kona 

 reef as compared, for example, with the Gulf 

 of California, may bear on the relative scarcity 

 in Kona of prey fishes in large schools. Reason 

 for the relative absence of such fishes in Kona 

 was discussed above. The many schooling prey 

 in the Gulf of California was held to be a major 

 factor in the occurrence there of so many large 

 piscivores (Hobson, 1968). For whatever rea- 

 son these faunal elements are missing in Kona, 

 the question remains: if the threat from pred- 

 ators strongly influences the well-defined se- 

 quence in which fishes in the Gulf seek and leave 

 cover during the day-night transition, how is it 

 that reef fishes in Kona show essentially the same 

 pattern when the threat of large piscivores to 

 them is comparatively mild? 



We have been focusing on the situations in 

 Kona and the Gulf of California, when in fact 

 it seems that related fishes on most, if not all, 

 tropical reefs, show similar behavior during twi- 

 light. Data are lacking for most areas, so this 

 conclusion remains tentative. Nevertheless, not 

 only do related fishes common to Kona and the 

 Gulf of California behave similarly, as described 

 above, but limited data from other seas, includ- 

 ing the tropical Atlantic (Starck and Davis, 

 1966) show essentially the same pattern. The 

 similar pattern of events in these widely sep- 

 arated areas indicates a long evolution that in 

 any one locality transcends the existing situa- 

 tion, even the existing species. 



One would expect predator-prey relations 

 among fishes to have responded similarly 

 throughout time to certain basic, unchanging 

 phenomena. Among others, these phenomena 

 would have included the characteristics of light 

 underwater and the efl^ect of these character- 

 istics on visually feeding predators at diff"erent 

 times of the day, as well as the probable diffi- 

 culties such predators experience in choosing a 

 target from among the many confronting them 

 in a fish school (Allen, 1920a, 1920b; Hobson, 

 1968 ; and others) . With this common heritage, 

 it would be surprising if behavior patterns re- 

 lating to predator-prey interactions did not show 

 today deep-rooted parallels on even widely sep- 

 arated reefs. I suggest, therefore, that the be- 

 havior of tropical reef fishes during twilight in 

 all seas, including Kona, reflects the impact of 

 a historic threat from predators. 



CONCLUSIONS 



1. At least most of the reef fishes at Kona, 

 Hawaii, behave diflferently in daylight than they 

 do at night, and during twilight their actions 

 express a transition between these two modes 

 of behavior. Most of these fishes are either di- 

 urnal or nocturnal, with the diurnal species rel- 

 atively inactive at night and the nocturnal spe- 

 cies relatively inactive in daylight. 



2. During their period of major activity, the 

 actions of these fishes are dominated by behavior 

 relating to their own feeding, and often this 

 takes them to specific feeding grounds. During 

 their inactive periods their behavior relates 

 strongly to their own security, and usually they 

 take shelter. Consequently, the actions of these 

 fishes during twilight relate primarily to moving 

 between feeding locations and shelter locations. 



3. The twilight movements of many of these 

 fishes involve well-defined migrations. A) Some 

 members of certain nocturnal species migrate 

 extensively between shelter locations on the reef 

 and feeding grounds ofl'shore; species in this 

 category include Myripy-istis spp, and Priacan- 

 thiis cruentattis. B) Some members of certain 

 other species migrate between shelter locations 

 on one part of the reef and feeding grounds on 

 another; species in this category are mostly di- 



738 



