44 
Fishery Bulletin 99(1 ) 
A o 
20 
40 
I 
0 12 3 4 
Vessel speed (knots) 
5 
Time (hours) 
0:00 4:00 8:00 12:00 1600 20:00 
10 
6 !L 
w 
1 
- 4 
2 
C 0 
20 T I 
<1 " " 
1 40 1 
* :|i i i h 1 
100 4 r t T , 
0 1 2 3 4 5 
Vessel speed (knots) 
Figure 4 
Relationships between swimming depths of the fish and the 
speed of the tracking vessel during associations. (A) Mean 
values (±SD) corresponding to each association: Tl, T2, T3 for 
tunas 1, 2, 3; dl, d2, representing day times (white dots) and 
nl, n2 representing night times (black dots) of the first two 
24-h cycles of the track of tuna 3. The two other graphs repre- 
sent details of the association between tuna 3 and the track- 
ing vessel during the third day (22 April 1996) where many 
changes in the vessel speeds occurred: (B) vertical movements 
and vessel speeds in relation to time; and (C) depth versus 
vessel speed (mean value ±SD). 
it was captured (approximately 150 m) and immedi- 
ately became associated with the tracking vessel. It re- 
mained at a depth between 80 m and 150 m while the 
vessel drifted close to the FAD. When the ship left the 
FAD (FAD1) at 16:20 to go to an other FAD (FAD2 
South of Moorea Island), the fish moved closer to the 
surface and followed in a similar manner to that ex- 
hibited by yellowfin tuna 1 and 2. It remained within 
a school of other yellowfin tuna (individual size rang- 
ing from 20 to 50 kg) that could be seen from the ship, 
swimming just below the surface. During the second 
day (21 April), the fish remained associated with the 
ship as it drifted close to FAD2. Mean swimming depth 
was 75 m. During the second night (21-22 April), the 
vessel moved from FAD2 to FAD3, southwest of Maiao 
Island, and the fish followed, swimming close to the 
surface as during the first night. When the vessel and 
the fish arrived at FAD3 in the morning of 22 April, the 
fish left the surface and came under the ship. Around 
07:45, the vessel left the FAD to shelter close to Maiao 
Island to find better sea conditions, with the tagged yel- 
lowfin tuna and the school following closely. The fish 
remained associated when the vessel returned to the 
FAD at 12:30. In the afternoon, the vessel drifted close 
to the FAD until 19:00, after which it made rapid ac- 
celerations away from the FAD to break off the asso- 
ciation. The strategy was successful, and the fish re- 
mained associated with the FAD. During the fourth day 
(23 April), trawl and acoustic survey operations were 
conducted away from the FAD. The fish did not associ- 
ate with the vessel during this period, but its presence 
at the FAD was regularly observed. Attempts to re- 
associate the fish with the tracking vessel were not suc- 
cessful because the fish returned to the FAD when the 
ship moved 0.5 nmi from the FAD. The fish remained 
associated with the FAD until operations terminated. 
Figure 4 shows the relationships between swimming 
depths and the speed of the vessel during all observed 
associations. Shallower swimming depths were observed 
when the vessel was moving at higher speeds, both dur- 
ing daytime (tuna 1 and 2) and nighttime (tuna 3) as- 
sociations. Figure 4 also shows in detail the response 
of tuna 3 to rapid changes in vessel speed during the 
third day (22 April), when the vessel was moving be- 
tween FAD3 and the Maiao Island, or drifting close to 
the FAD or to the island. This day represents an exam- 
ple of frequent changes in vessel speed and correspond- 
ing changes in swimming depths of the fish. 
Discussion 
Individual yellowfin tuna have been documented asso- 
ciating with tracking vessels in the Indian Ocean (a 
108-cm yellowfin tuna, Cayre et ah, 1996) and near the 
main Hawaiian Islands (a 167-cm yellowfin tuna, Brill 
et al., 1999). However, these authors merely noted the 
occurrence of the associations without providing fur- 
ther analyses or comments on this striking behavior. It 
is noteworthy that all these fish were yellowfin tuna. It 
