FISHERY BULLETIN: VOL. 77, NO. 4 



surface as a wall through which the prey cannot 

 escape. This function has also been suggested by 

 Norris and Dohl (in press), and shown to be carried 

 out by some large predatory fishes by Major (1976; 

 cited by Norris and Dohl in press). 



Noisy leaps, which started at some point before 

 the fish school appeared, continue throughout sur- 

 face feeding. We gained the subjective impression 

 that these leaps occurred on the periphery of feed- 

 ing bouts. This may be because breaching directly 

 into the fish school would certainly not be of ad- 

 vantage in keeping it tight against the water sur- 

 face. As well, it may serve to keep fish from escap- 

 ing, and thus may be an acoustic or vibration 

 "netting" effect. We often saw dolphins tailslap- 

 ping rapidly (2-3 slaps/s) while moving in a tight 

 circle around feeding bouts, and this action may 

 further serve to keep fish from escaping (it has 

 been described by Norris and Dohl (1980) and 

 labeled "motorboating"). 



Besides the function of recruiting nearby groups 

 to the feeding bout either purposefully or inciden- 

 tal to keeping fish from escaping, there is a third 

 possibility. The splashes of noisy leaps create an 

 underwater omnidirectional sound which may ac- 

 tually serve to frighten fish and cause them to 

 school more tightly. Although work has been done 

 on schooling relative to pressure waves (Bobbi 

 Low*^) asfar as weknow, no studies exist on sounds 

 and fright reactions in schooling fish. 



While the feeding bout continues, clean leaps 

 and humpings continue as well, and dolphins still 

 dive steeply. They also come up at a sharp angle, 

 and individuals move rapidly through the 

 bunched fish school, appearing at the other side of 

 the school with several fish in their mouths. They 

 then dive steeply again, and usually resume "at- 

 tacks" on the fish school from below. The dolphins 

 may stay below the fish much of the time to keep 

 the school from escaping downward, and possibly 

 to herd other fish to the surface to continue or 

 prolong feeding. This is not the first indication of 

 apparent cooperative herding and feeding in dol- 

 phins. It appears that many different species coop- 

 erate in herding, and it has been described for 

 representatives of the genera Orcinus. Tursiops, 

 Sousa, Phocoena, Delphinus, and others (see Nor- 

 ris and Dohl in press). Many terrestrial predators 

 do so as well (Wilson 1975). 



'Bobbi Low, professor, University of Michigan, Ann Arbor, MI 

 48109, pers commun. 1976. 



Acrobatic noisy leaps are most often seen during 

 and after feeding. These may herd fish and recruit 

 nearby groups, but they appear to require much 

 energy and coordination which seems unneces- 

 sary just to make noise. We believe with Norris 

 and Dohl (1980) that they may serve a "social 

 facilitation" function, signaling a high activity 

 level as individuals reaffirm and strengthen social 

 and possibly sexual bonds. Saayman and Tayler 

 (1973, 1979) describe similar high activity 

 levels in Sousa sp. when two or more groups meet, 

 and provide a similar assessment. We suggest that 

 individual animals have taken care of the basic 

 requirement of feeding and are now prepared to 

 spend time socializing and "playing." 



After feeding, dusky dolphins are more willing 

 to associate with boats, human swimmers, whales, 

 sea lions, and inanimate objects such as kelp. This 

 may be an outgrowth of the high level of social 

 activity at that time. Although we also saw much 

 apparent mating after feeding, we were not able to 

 compare it with amount of mating in small, 

 nonsurface-feeding groups. 



When many small groups coalesced to form a 

 large one, did the smaller units remain intact or 

 was movement of dolphins throughout the large 

 group "random"? We saw individuals which had 

 been spaghetti-tagged in a small group traveling 

 together within a feeding bout a few days after 

 tagging (Wiirsig in press), and thus have some 

 indication that the small group remained intact. 

 This agrees with data by Norris and Dohl ( 1980) 

 on Hawaiian spinner dolphins. They found that 

 there is fluidity in schools, but that small groups of 

 4-10 animals may be the only units with longer 

 term continuity. We have no long-term informa- 

 tion on group stability. However, studies of other 

 dolphins suggest that the small-unit group com- 

 position is constantly changing (Saayman and 

 Tayler in press, humpback dolphins; Shane 1977; 

 Wiirsig 1978; Wells et al. in press, bottlenose dol- 

 phins). This flexibility in small-group composition 

 at least superficially resembles chimpanzee group 

 structure, and Saayman and Tayler (1979) and 

 one of us (Wiirsig 1978) independently speculated 

 that the similarity comes from feeding on unpre- 

 dictable and patchy food distribution (see also 

 Nishida 1968). If a similar group structure is 

 found in dusky dolphins, it might be possible that 

 individuals move randomly throughout the large 

 after-feeding group, and that the entire group of 

 up to 300 animals forms the more stable breeding 

 unit or population. 



886 



