FISHERY BULLETIN: VOL. 80. NO. 2 



School 5 behaved quite differently from the 

 others. Both relative movement, time of radar 

 fixes, and swimming speed vectors are shown for 

 this school (Fig. 4). The swimming speed vectors, 

 shown as arrows attached to the relative motion 

 line between various time and distance intervals, 

 are drawn proportional to the calculated 

 swimming speeds (Table 3). 



School 5 was probably feeding when the bird 

 flock associated with it was first detected on the 

 ship's radar at a distance of 5.8 nmi. The distance 

 and bearing plots of the birds indicated erratic 

 movement. Later, in the tracking-by-helicopter 

 phase, the first two ranges and bearings showed 

 the school moving at only 2.6 kn. The inferred 

 feeding behavior from this is consistent with the 

 feeding behavior described by Norris et al. 

 (1978), as well as other observations by us in the 

 eastern Pacific. At a closer range of about 3.3 

 nmi from the ship, the school's behavior changed 

 radically as it altered course by 97° to the right 

 and increased its speed to 8.4 kn, turning on a 

 course that would have taken it 2.0-2.5 nmi 

 abeam of the passing ship. Between 2.3 and 3.0 

 nmi the school again shifted course, this time by 

 70° to the left, and increased its speed to 9.4 kn. 

 When this school reached a point about 0.5 nmi 

 from the track line and 2.3 nmi from the ship, its 

 behavior changed again. Individuals and 

 subgroups within the school began swimming in 

 many different directions, making large changes 

 in course heading. Suddenly the main body of the 

 school turned nearly 180° and swam toward the 

 ship at high speed (~9 kn). After closing to 

 within 1 nmi of the ship, the school reversed itself 

 again, and thereafter swam rapidly away from 

 the vessel. This type of "error" in choice of 

 avoidance heading was only seen in schools 4 and 

 5, which were both relatively large schools (300- 

 350 individuals estimated). 



School Speed 



While avoiding the ship, the speeds of the first 

 seven schools varied between 2.5 and 13.1 kn, 

 with average speeds between 5.1 and 8.8 kn 

 (Table 3). The eighth school was too close to the 

 ship for ranging measurements by radar. There 

 was no apparent difference in swimming speeds 

 among the three species observed. Substantial 

 variation in speed occurred in all seven schools. 

 Schools 1, 2, 3, and 4 swam at speeds that aver- 

 aged between 5 and 7 kn. Schools 6 and 7 had the 



Table 3.— School swimming speed vectors. 



'Range and bearing of school from ship at times appropriate to the 

 speed calculation If notable, behavior at these and other times are re- 

 ported in Table 1. 



2 Speed vectors pertain to time intervals ending at times listed unless 

 otherwise indicated Calculations are from relative or absolute plots in- 

 volving ship motion. 



3 Mean school speed refers to times when school is responding to ship 



4 Time interval for this calculation. 



5 This school was actually sighted at 0953. but measurements did not 

 begin until 1001 



376 



