MAJOR: PREDATOR-PREY INTERACTIONS IN FISHES 



attacks made by predatory fish. This protective 

 function has also been observed for other school- 

 ing prey species (Radakov 1958, 1973; Neill and 

 Cullen 1974). The chance that a predator has of 

 singling out a specific individual silverside are 

 greatly reduced if schools are formed. This 

 appears to be especially true when the prey are 

 polarized towards one another and move close 

 together through coordinated maneuvers. In the 

 field, when predators were not in the immediate 

 vicinity of silverside schools, individual silver- 

 sides became relatively motionless and randomly 

 oriented towards one another, darting out from 

 schools presumably to feed. When individual 

 silversides presumably became exposed and/or 

 appeared to be accessible to one or more nearby 

 predators, the predators approached or attacked. 

 If the predator's approach was slow, the individual 

 silversides became polarized, the school maneu- 

 vering evasively. If a predator's approach was 

 sudden or rapid, individual silversides jumped out 

 of the water one or more times to evade. Both 

 schooling and jumping presumably decrease the 

 time a predator had to align itself with a specific 

 individual prey. In addition, a jumping silverside 

 often landed in the midst of its own, or that of 

 another nearby, school, presumably disappearing 

 from the predator's field of vision and/or path of 

 swimming. The formation of large schools com- 

 posed of many hundreds or thousands of indi- 

 viduals, especially a number of such schools 

 relatively close to one another, appeared to 

 increase an individual silverside's chance of 

 escape when jumping. 



The movement of silversides into the shallow 

 water over reefs, and their location near and 

 under Lilipuna Pier and heavy overhanging 

 vegetation and along the sides of the lagoon, may 

 be additional means, besides schooling, of reduc- 

 ing predation. In the shallow water near the pier, 

 the most common vertical attacking predators 

 were lizardfish. In deeper water in the lagoon and 

 near the pier, jacks and leatherjackets also 

 attacked vertically. Horizontal stalking and at- 

 tacking predators, such as barracuda and needle- 

 fish, occurred in both deep and shallow water. The 

 depth of water over the reefs may have been less 

 than sufficient for some of the vertical attacking 

 species to maneuver and approach schools of 

 silversides undetected. The occurrence of silver- 

 sides near structures and along the sides of the 

 lagoon may have also limited the maneuver- 



ability and avenues of approach for all species of 

 predators. 



DISCUSSION 



The interactions between silversides and their 

 predators in relation to solar phenomena are 

 almost identical in pattern and time to those given 

 by Hobson (1968, 1972) for the interactions of 

 Hurengula thrissina and their predator Mycter- 

 operca rosacea in the Gulf of California. Hobson 

 and Chess's (1973) study of the arrival and 

 departure of Pranesus pinguis to and from reefs at 

 Majuro Atoll in the Marshall Islands also showed 

 school movement related to specific times during 

 twilight. However, only a few predatory attacks 

 were observed at Majuro Atoll. Comparisons of 

 lunar and tidal changes during the studies in 

 Kaneohe Bay and Majuro Atoll and Baja Califor- 

 nia seem to indicate a relatively minor influence 

 on the crepuscular behavior of schools. 



Hobson (1968, 1972, 1973), Collette and Talbot 

 (1972), and Domm and Domm (1973) have 

 demonstrated that there is relatively little activ- 

 ity amongst most coral reef fishes during a specific 

 segment of the twilight period. In the morning, 

 nocturnally active reef fish leave the open water 

 column to hide in the coral reef approximately 30 

 min before sunrise (Hobson 1972). Diurnal species 

 do not reoccupy the water column until approx- 

 imately 12-16 min prior to sunrise. It is exactly 

 between the above times, the "quiet period," as 

 defined by Hobson (1972), that peak surface 

 predator-prey activity and school formation takes 

 place in Kaneohe Bay, just as it does in the Gulf of 

 California (Hobson 1968, 1972), and possibly 

 Majuro Atoll (Hobson and Chess 1973). The 

 pattern is reversed during evening twilight 

 (Hobson 1972). Diurnal reef species evacuate the 

 water column approximately 6-22 min after 

 sunset. Nocturnal species then reoccupy the water 

 column about 14-34 min after sunset. Again, 

 surface predator-prey interactions peak and 

 schools break up in Kaneohe Bay during the time 

 that would be comparable with the evening quiet 

 period in other parts of the world. 



The combined observations of reef fishes in the 

 Virgin Islands (Collette and Talbot 1972), the 

 Great Barrier Reef, Australia (Domm and Domm 

 1973), Hawaii (Hobson 1972), and the Gulf of 

 California (Hobson 1968) indicate nearly identi- 

 cal time relationships of behavioral events during 



423 



