FISHERY BULLETIN: VOL. 70, NO. 1 



DISCUSSION 



These results demonstrate that blockage of 

 hydrodynamic contact between fish in a school 

 resulted in a significant alteration in side-to-side 

 spacing and a changed ratio of diagonal to abeam 

 positional orientation. Of the various sensory 

 factors of importance in the changes observed, 

 it appears that the loss of acoustico-lateralis cues 

 from neighboring fish played a key role. Some 

 of the alternative explanations are discussed and 

 dismissed below. We did not do any acoustic 

 calibrations; the cement channels and plastic 

 barriers may have caused some resonance of fish 

 swimming sounds, and these may act as orienta- 

 tion cues. But this would assume that the sounds 

 were above the auditory thresholds for these fish, 

 and that the sound can be directionally localized ; 

 the latter supposition we can hardly make, as yet, 

 even for other fish whose hearing ability has been 

 extensively studied, with the exception perhaps 

 of goldfish (Moulton and Dixon, 1967). We 

 know nothing about hearing in kawakawa; the 

 yellowfin, Thunnus albacares, has moderately 

 good acoustic sensitivity, and also has a swim- 

 bladder, an organ that may be helpful for sound 

 pressure to reach the ear (Iversen, 1967). Ka- 

 wakawa lack a swimbladder; we therefore be-^ 

 lieve it unlikely that sonic pressure is as im- 

 portant in the lives of these fish as is hydrody- 

 namic particle motion. The latter was blocked 

 by our partitions, so that mate-generated cues 

 to the lateral line mechanoreceptors were ob- 

 structed. 



There are several hydrodynamic factors to be 

 considered: while it is true that the partitions 

 changed the hydrodynamic field around the fish, 

 and that the solid Plexiglas could possibly cue 

 him to move farther away and thus increase the 

 fish-to-fish spacing, we do not believe this to rep- 

 resent the primary dynamics involved. The in- 

 creased spacing appeared from our data to rep- 

 resent primarily the fish-to-fish orientation 

 (Figure 4); this orientation resulted in fish 

 moving close to the partition just as often as 

 they moved away from it, while tracking their 

 mates. At all times the fish keep a certain min- 

 imal distance of 7 to 8 cm from the barrier, so 

 that we can say that there is some "orientation" 



to the partition. Superimposed on this is the 

 fish-to-fish orientation, and this exerts the major 

 control of the side-to-side spacing. 



Hydrodynamically, when two fish of about the 

 same size are swimming side by side (abeam 

 position) at about the same speed, there would 

 be virtually no centrally located vortices between 

 them. This is because the vortices generated by 

 each fish trail slightly behind him (Rosen, 1959) . 

 The flow patterns at the center would, under 

 these conditions, not be very useful as cues. For 

 the fish to get maximal mate-generated hydro- 

 dynamic cues it therefore appears that the di- 

 agonal positional orientation is preferable. This 

 preference for "diagonal position to each other 

 over positions in front, behind, or directly a- 

 beam" was pointed out by Cullen, Shaw, and 

 Baldwin (1965) in their three-dimensional anal- 

 ysis of fish school geometry. Van 01st and 

 Hunter (1970) also called attention to the fact 

 that "the minimum possible lateral space be- 

 tween neighbors to the side would be lower if 

 fish consistently occupied diagonal positions than 

 if they were perfectly aligned, because the areas 

 of tail movement would not overlap." In re- 

 lationship to our experimental results, the par- 

 titions blocked the mate-generated vortices so 

 there was no longer any "advantage" to the di- 

 agonal position. The "advantage" factor, it 

 should be emphasized, is not only for hydrody- 

 ftamic cues from schoolmates, but also more im- 

 portantly, to derive maximal locomotor efficiency 

 while travelling in a school. The fish can best 

 use the energy from mate-generated vortices if 

 he is positioned within the influence of the vor- 

 tex, and this is usually best at primarily right 

 angles and a little behind the course of the fish 

 (Breder, 1965), the diagonal position. 



Since the fish, in the presence of the partitions, 

 did not completely abandon the diagonal posi- 

 tioning while increasing the abeam orientation, 

 the question comes up of whether or not vision 

 is equally good from both positions ? Tuna vision 

 has been studied to a limited extent; for example 

 Nakamura (1968) studied visual acuity in kawa- 

 kawa by testing the fish's response to various 

 targets. This involved the temporal region of 

 the retina as the fish swam forward towards the 

 target. A different region of the retina is no 



202 



