688 



Fishery Bulletin 90(4), 1992 



But when the prey are distributed in rare patches 

 and the dolphins are more adept than the tuna at 

 locating these patches, then tuna could benefit from 

 associating with dolphins because the fish could en- 

 counter food more often than if they were not asso- 

 ciated. This will be true regardless of the density of 

 the prey patch. 



It is never the case that dolphins benefit energetically 

 from depending entirely on tuna for finding prey, 

 because dolphin forage requirements are so much 

 higher than tuna requirements. 



These conclusions lead to the hypothesis that tuna- 

 dolphin associations should be more prevalent in areas 

 where oceanic conditions encourage strong clumping 

 of prey, and less prevalent when conditions encourage 

 a more homogeneous distribution of prey. I am current- 

 ly exploring, with a simulation model of tuna, the 

 movements of dolphin and prey in response to envi- 

 ronmental characteristics of the ETP (work in pro- 

 gress). Further studies correlating oceanic environmen- 

 tal characteristics with catches of various size-classes 

 of tuna are planned but not yet underway. If the sug- 

 gestions described above are borne out, it may be pos- 

 sible to identify areas of the ETP where large yellowfin 

 tuna could be captured without having to rely on 

 dolphin-associated fishing. 



Caveats 



This study assumes that average size of dolphin schools 

 remains constant at about 200 animals. This is the 

 average school size for spotted dolphins observed dur- 

 ing dolphin survey research cruises in the ETP. In fact, 

 neither school size nor school composition are constant. 

 Observers on both research and commercial vessels 

 report school sizes ranging from a few animals to many 

 hundreds. Scott (1991) reports diel changes in sizes of 

 schools sighted by tuna fishermen in the ETP. 



However, these inconsistencies may not significant- 

 ly affect the implications of the energetics estimates 

 presented here. Average sizes of dolphin schools cap- 

 tured with tuna in the ETP are considerably larger 

 (400-600 animals) than the average school size ob- 

 served during research surveys because the fishermen 

 preferentially search and capture large schools of 

 dolphins, which tend to carry more tuna. Estimates 

 concerning the relative importance of tuna and dolphins 

 to energetics of the association are probably reasonably 

 similar for both large and small associations, because 

 in both cases the proportions of tuna and dolphins tend 

 to be similar (i.e., as the number of dolphins increases, 

 in general the number of associated tuna increases). 

 The study of diel differences (Scott 1991) shows that 

 school sizes of dolphins sighted in association with tuna 

 vary from a morning low to a late-afternoon high, but 



the change is relatively small, from ~450 to ~600 

 animals on average. 



Other exptanations for the bond 



Other hypotheses have been proposed to explain the 

 tuna-dolphin association. The two most-often suggested 

 are the possibility that tuna perceive dolphin schools 

 as FADs (fish aggregating devices) or as protection 

 from sharks. Both of these factors may well contribute 

 to the strength of the bond; neither precludes the 

 energetics results discussed above. 



The propensity for fish to collect around floating ob- 

 jects is well known, although the reasons are not yet 

 understood. Presumably, floating objects provide a 

 reference point for the aggregating tuna and in some 

 way increase foraging success, perhaps by concen- 

 trating prey items or by tracking convergence areas 

 where prey densities may be higher than elsewhere. 



The FAD hypothesis has merit for the sizes of tuna 

 actually found with dolphins in the ETP, for two 

 reasons in particular. First, associating with dolphins 

 may increase foraging success for the associated tuna 

 because both tuna and dolphins are apparently seek- 

 ing the same prey and dolphins may be more adept at 

 finding it. Thus, tuna are associating with a FAD that 

 does not simply attract appropriate prey passively, but 

 actively searches and finds it. Second, tuna are required 

 to swim constantly in order to ventilate their gills. It 

 appears convenient that the average observed speed 

 of dolphin schools is also the optimum speed of the sizes 

 of tuna usually found associated with these schools. 

 Rather than circling a stationary FAD, tuna associated 

 with dolphin schools will cover a much larger area while 

 moving at their most efficient cruising speed, and will 

 cover that area in the presence of a sentient foraging 

 FAD. 



The shark protection hypothesis derives from a com- 

 mon perception that dolphins actively protect their 

 young by driving sharks from their vicinity. If this is 

 so, tuna associating with dolphins may be associating 

 with the best of all possible FADs; a floating object that 

 moves at the tuna's optimal speed, moves in search of 

 the same prey the tuna would like to find, is probably 

 at least as adept as the tuna at finding that preferred 

 prey, and which provides protection against, rather 

 than increased risk of, predation (FADs of course con- 

 centrate not only fish, but also their predators). 



Both the FAD and shark hypotheses assume that 

 tuna follow dolphins. Not all hypotheses assume that 

 tuna are the followers. Au and Pitman (1986) and Au 

 (1991) suggest, for example, that dolphins follow tuna 

 in order to take advantage of tuna foraging in conjunc- 

 tion with bird flocks. This would be an advantage for 

 dolphins during the actual feeding event. However, it 



