Brill et a\: Horizontal and vertical movements of iiivenlle Thunnw, thynniK 



157 



positions were obtained by using a GPS satellite receiver 

 and were recorded on a second laptop computer every 

 minute. The tracking vessel's position was assumed to be 

 the same as that of the fish. Sea surface temperature and 

 bottom depth were recorded manually every 15 minutes 

 by using a hull-mounted electronic temperature sensor 

 and color fathometer, respectively. Depth-temperature 

 profiles were taken approximately every four hours with 

 a Sippican (Marion, MA) portable XBT system (model 

 MK12). 



Aggregate time-at-depth and time-at-temperature dis- 

 tributions were calculated fromlO-m and 1°C bins (respec- 

 tively), as described by Holland et al. (1990). These data 

 were subsequently expressed as a fraction of the total time 

 each fish was followed, and the fractional data bins were 

 averaged across all fish. Speed over ground (henceforth re- 

 ferred to simply as "speed") was calculated by assuming 

 that the fish moved in a straight line between successive 

 geographic locations. 



Sea surface temperature (SST) data were recorded by 

 the advanced very high resolution radiometer (AVHRR) 

 carried onboard the NOAA-14 polar orbiting operational 

 environmental satellite. High resolution picture transmis- 

 sion (HRPT) data were obtained from the National Coast 

 Watch Active Access System at the National Oceano- 

 graphic Data Center and had a spatial resolution of 1.25 x 

 1.25 km pixels. Ocean color data were recorded by the 

 Sea-viewing Wide Field-of-view Sensor (SeaWiFS) car- 

 ried onboard the Orbview-2 spacecraft (Orbimage, Inc., 

 Dulles. VA). The level-2 global area coverage (GAC) data 

 were obtained from the NASA Goddard Space Flight Cen- 

 ter's Distributed Active Archive Center. These 4-km reso- 

 lution data sets included chlorophyll-a surface concentra- 

 tion and the diffuse attenuation coefficient at 490 nmi, 

 m vacuo. We calculated the occupancy of waters with 

 specific chlorophyll-a concentrations and light attenua- 

 tions from values corresponding to and coincident with 

 the tracks of fishes derived from satellite images. These 

 data were subsequently expressed as a fraction of the to- 

 tal number of observations for each fish, and the fraction- 

 al data bins were averaged across all fish. For illustrative 

 purposes, we also generated composite images using data 



from the 21-day period over which all tracking operations 

 were conduced. 



Results 



The bottom topography in the areas where the fish were 

 tracked is generally featureless, except for small areas 

 where the vertical relief is approximately 2 m above the 

 surroundings. Local fishermen have named these features 

 (Fig. IB and subsequent figures), and the names used in 

 this study are taken from local fishing charts. 



Size of fish, starting and ending dates of tracks, dura- 

 tion of tracks, distances covered, distance between start- 

 ing and ending points, and mean (±SEM) swimming speed 

 offish are listed in Table 1. With the exception offish num- 

 ber 4 (referred to simply as "fish 4"), individuals tended 

 to follow highly irregular courses that often repeatedly 

 covered the same areas (Fig. IB). The mean distance be- 

 tween starting and ending points for all fish was only 

 ll'7f (range: 4— 25'/f ) of the total distance covered (Table 1). 

 From tracking studies of yellowfin tuna [Thitnniis alba- 

 cares) in the Pacific, Dagorn et al. (2000a) concluded that 

 such frequent directional changes might be characteristic 

 of foraging behavior. The frequency of observed swimming 

 speeds is shown in Figure 2. Although all fish reached 

 maximum speeds of =7 knots for brief periods, over 90% of 

 the observed speeds were less than 3.6 knots. 



Horizontal movements 



Fish 1 was captured and released at 1340 h, approximately 

 1.8 km (1 nmi) west of the "26 Mile Hill" (Fig. 3). It pro- 

 ceeded on a southerly course for about 33 h, a direction that 

 carried it over the "Hot Dog" and "Southeast Lumps." After 

 sunset on the second day, the fish reversed its course and 

 eventually recrossed both features. The fish was approxi- 

 mately 5.6 km (3 nmi) south of the "Southeast Lumps," and 

 moving south, when the track was terminated at 1300 h. 



Fish 2 was captured approximately 5.6 km (3 nmi) south 

 of the "Southeast Lumps" (Fig. 3) at 1547 h, adjacent to 

 where the track of fish 1 was completed four days earlier. It 



