494 



Fishery Bulletin 88(3), 1990 



the horizontal distance between the 40-fathom and 

 100-fathom (187-m) isobaths is measured in tens of 

 meters. To observe the behavior of non FAD-associated 

 tuna, three yeliowfin tuna were tracked that were 

 caught on or near the 40-fathom isobath around Oahu, 

 Hawaii. 



The influence of FADs on the vertical movements of 

 these species was investigated by comparing the depth 

 distributions of the two species in on-FAD (within 500 

 m of a FAD) and off-FAD (beyond 500 m of a FAD) 

 situations. The depth distributions were analyzed with 

 respect to ambient ocean temperatures, as measured 

 by expendable bathythermographs. 



Methods 



The horizontal and vertical movements of individual 

 fish were monitored by pressure-sensitive, 50 KHz 

 ultrasonic transmitters (Vemco, Halifax, Nova Scotia, 

 Canada). Fish caught on the tracking vessel by troll- 

 ing and pole-and-line techniques were fitted with trans- 

 mitters by passing two nylon "tie-wraps" through the 

 dorsal pterygiophores and trunk musculature adjacent 

 to the second dorsal fin (Fig. 1). The fish were out of 

 the water for approximately 1 minute during this pro- 

 cedure. The fish were then released and followed using 

 a directional hydrophone mounted on the tracking boat. 

 Whenever possible, a distance of approximately 200 m 

 was maintained between the fish and the boat. Geo- 

 graphic position was determined every 15 minutes by 

 a combination of Loran-C, radar, bathymetric, and 

 visual fixes. These techniques evolved from pioneer- 

 ing work in tuna tracking by Yuen (1970), Laurs et al. 

 (1977), Carey and Olson (1982), and Carey (1983). A 

 summary of previous ultrasonic tracks of tuna has been 

 compiled by Hunter et al. (1986). A detailed account 

 of the methods used in the current study has been 

 published previously (Holland et al. 1985, Bayliff and 

 Holland 1986). 



The transmitters were equipped with pressure sen- 

 sors, which modulated the rate of pulse transmission 

 in response to changes in water pressure (depth). Thus, 

 vertical movements of the fish were determined by 

 measuring the time between signal pulses. These pulses 

 were recorded on audiocassettes for analysis ashore. 

 Expendable bathythermographs deployed approx- 

 imately every 3 hours provided ocean temperature pro- 

 files, which were superimposed on the vertical move- 

 ment plots. Pooled time-at-depth and time-at- 

 temperature histograms were constructed from all 



Reference to trade names does not imply endorsement by the 

 National marine Fisheries Service, NOAA. 



Figure 1 



Transmitters were attached to yeliowfin and bigeye tuna with two 

 nylon straps inserted through the dorsal musculature and pterygio- 

 phores associated with the second dorsal fin (from Holland et al. 

 1985). 



tracks combined, using 10-m and 1°C bins averaged 

 every 10 minutes. The data were analyzed for daytime 

 and nighttime depth distributions. Differences in the 

 temperature and depth distributions for day-versus- 

 night and on-FAD versus off-FAD were analysed by 

 factorial analysis of variance (SAS General Linear 

 Models Procedure, SAS 1985). Portions of tracks oc- 

 curring on-FAD or near the reef dropoff were not in- 

 cluded in these calculations. Thus, the depth histograms 

 and overall averages are compiled from movements 

 that were not constrained by bottom topography or in- 

 fluenced by the presence of floating objects. 



The tracking techniques used in this study are not 

 capable of detecting small-scale changes in the horizon- 

 tal swimming direction of the fish. Consequently, sus- 

 tained swimming speeds have been calculated only 

 from sections of tracks in which fish demonstrated pro- 

 longed straight-line movements. Swimming speeds 

 were calculated as distance traveled per unit time and 

 in terms of body lengths (fork length, FL) per second. 

 Swimming speeds were calculated for each complete 

 hour of "staight" running; these hourly rates were 

 averaged to yield a sustained swimming speed for the 

 relevant section of each track. Vertical movements 

 were not included in these calculations. 



Daytime and nighttime time-at-temperature histo- 

 grams were generated for both yeliowfin and bigeye 

 tuna. Because of differences in absolute ocean-surface 

 temperatures between different tracks, and because of 



