386 



Fishery Bulletin 105(3) 



Time (local) 



1:55 12:00 12:05 12:11 12:16 12:21 1227 12 32 12.38 12:43 12:48 12:54 12:59 13:05 

 ^^—t — . ' ' 



■B J"o  



Q. 

 0) 



Q 



400 



TT 



3 

 "a 



(C 



III! 5 

 c 

 5 



us O 



Figure 5 



Depth and temperature data for a skipjack tuna {Katsuwonus pelamis) (tag 

 no. 4930) exhibiting deep-diving behavior. 



studies on the simultaneous behavior of skipjack, big- 

 eye, and yellowfin (Thunniis albacares) tunas associated 

 with drifting fish aggregation devices in the central Pa- 

 cific Ocean also revealed that skipjack tuna swimming 

 depths are predominantly shallower than the thermo- 

 cline depth, about 115 m, throughout the day and night, 

 but average nighttime depths are shallower than those 

 during the daytime (Matsumoto et al.^). 



The vertical movement patterns observed in this study 

 for unassociated behavior of skipjack tuna, seen mostly 

 as relatively high frequency repetitive bounce diving 

 between dawn and dusk, was unexpected on account of 

 the data from previously published studies (Matsumoto 

 et al., 1984). There are, however, most likely other ver- 

 tical movement patterns for skipjack tuna when unas- 

 sociated with floating objects, from that observed in 

 the present study. This pattern is probably dependent 

 on the spatial and temporal habitat as well as forage 

 availability. In the first published paper on acoustic 

 tracking of tunas (Yuen, 1970), ultrasonic transmit- 

 ters were used to study the behavior and movements 

 of skipjack tuna off Hawaii. One individual, tracked 

 for 8 days, undertook daily cyclical movements away 

 from and back to a bank at consistent times, and from 

 these movements Yuen (1970) first suggested that skip- 

 jack tuna can navigate and have a sense of time. The 

 method Yuen (1970) used was suitable only to conclude 

 that the fish remained close to the surface at night but 

 could be found at various depths during the day. In a 

 subsequent tracking study in Hawaiian waters, three 

 skipjack tuna tagged with ultrasonic transmitters spent 

 time between the surface (23.5°C) and 263 m (13.5°C) 

 during the day, but remained above 75 m (22°C) depth 

 at night (Dizon et al., 1978). Vertical positions of the 



skipjack tuna were determined only every 3 minutes, 

 but indicated many rapid vertical movements, includ- 

 ing some to depths below the thermocline. Although 

 the three tracked skipjack tuna spent 85% of the time 

 in water above 20°C (about 160 m), there were brief 

 dives to temperatures between 12° and 14°C. The data 

 retrieved from seven archival tags recovered from skip- 

 jack tuna released off northern Japan indicated that 

 their nighttime movements were normally between the 

 surface and depths of about 30 m, but during the day- 

 time they exhibited frequent dives to depths below the 

 thermocline and below ambient temperatures of 12°C; 

 the greatest depth was reported to be 267 m (Ogura^ ). 

 The lowest peritoneal cavity temperatures recorded in 

 that study were around 17°C, whereas in the present 

 study it was about 16°C. Skipjack tuna held in land- 

 based tanks in Hawaii subjected to gradually lowered 

 temperatures resulted in one mortality at 17°C, and 

 none survived at 15°C for more than a few hours (Dizon 

 et al., 1977). 



A relatively dense DSL was observed at night on the 

 echosounder during the time the skipjack tuna were 

 associated with the vessel in this study. Skipjack tuna 

 associated with a TAO mooring in this study area have 

 been documented to be feeding at night near the surface 

 on prey organisms of the DSL (Schaefer and Fuller, 

 2005). The repetitive bounce diving observed in the 

 present study for unassociated skipjack tuna during the 

 day apparently reflects foraging activity on prey organ- 

 isms of the diurnally migrating DSL (Kuznetsov et al., 

 1982). The profile in Figure 3A of repetitive bounce 

 diving shows a higher frequency of dives during the 

 first hour at around dawn when the DSL is descend- 

 ing and during the 1.5 h period preceding dusk during 



