always discrete or isolated units. Other quarters also 

 show large areas of above-average CPUEs within which 

 high-CPUE cells are grouped. 



Given the set of numerals denoting high-CPUE posi- 

 tions by quarters, one could show arrows pointing in 

 directions other than those shown, but such a change in 

 direction would be contrary to the seasonal movements 

 of the surface fishery (and also tag return data) in areas 

 where this is known. In Japanese waters, for example, 

 the skipjack tuna are known to enter the fishery from 

 the south and southwest, continue northeastward along 

 the southern coast of Japan and then move offshore in 

 an easterly direction with the progression of the seasons 

 from spring through late fall (Kimura 1949; Kasahara 

 1971; Kasahara et al. 1971); in the eastern Pacific, the 

 skipjack tuna are known to enter the northern fishery 

 off central Mexico in about the second quarter, move 

 northward in succeeding quarters, and presumably move 

 offshore and back to the central equatorial areas in late 

 faU (Fink and Bayliff 1970; Williams 1972). Although the 

 direction of seasonal movement of fish in the eastern 

 Pacific is opposite to that in the western Pacific, certain 

 factors that could determine the movement pattern are 

 common to both areas. First, the east- west component 

 of the movement of high-CPUE cells generally follows 

 that of the prevailing surface currents. Second, there is 

 a poleward shifting of the high-CPUE cells correspond- 

 ing to a similar shifting of the warm isotherms during 

 late spring to fall, and an equatorward shifting during 



winter and early spring. The net effect of surface 

 current flow and poleward shifting of skipjack tuna in 

 response to seasonal warming of the water thus would 

 result in roughly circular paths. 



On the basis of these observations, roughly circular 

 paths were drawn (Fig. 12) in the direction of the 

 prevailing currents (Fig. 13) to indicate the probable 

 movement of skipjack tuna in aU regions of the Pacific. 

 Thus, in the southern hemisphere, where the major 

 ocean currents flow counterclockwise around the South 

 Pacific Central Water mass, the movement of skipjack 

 tuna are shown in counterclockwise paths. In the north- 

 ern hemisphere, where the major currents flow clock- 

 wise, the paths of skipjack tuna movement are also 

 clockwise, except in the eastern Pacific, where current 

 flow around the North Pacific Equatorial Water mass is 

 counterclockwise. The correspondence of skipjack tuna 

 movement to current flow is believed to be more than 

 incidental, as hypothesized for albacore in the waters off 

 Japan and yellowfin tuna off eastern Australia by Naka- 

 mura (1969). 



The several paths depicted in the figure suggest that 

 skipjack tuna in the Pacific Ocean may be composed of 

 different groups (not necessarily separate subpopula- 

 tions), possibly seven in the northern and seven in the 

 southern hemispheres. For ease of discussion, these will 

 be referred to as environmentally defined "groups" of 

 skipjack tuna. A schematic diagram of the group move- 

 ments is presented in Figure 14. 



Figure 12.— Quarterly changes in position of high-CPUE areas in the Japanese tuna longline fishery, 1964-67. The shaded areas 

 denote above-average CPUE areas in the first quarter of 2 or more years. 



22 



