MATHER ET AL : TAGGED BLUEFIN TUNA 



Table 1. — Release groups used for analysis of migrations of tagged bluefin tuna. 



variance-covariance matrix. The determinant of 

 the recapture variance-covariance matrix is pro- 

 posed by Bayliff and Rothschild as an index of the 

 dispersal of the fish. When the distance of the 

 recoveries from one another is large, the deter- 

 minant is large. 



It should be noted that the vectors computed by 

 this method are not on a per-unit-effort basis so 

 that "migration patterns" reflect not only the ap- 

 parent movement of the fish, but also the distribu- 

 tion of fishing effort. In order to more fully under- 

 stand the nature of short-term movements, it will 

 be necessary to study in some detail the complex 

 problem of the distribution of bluefin tuna in the 

 northwest Atlantic. Preliminary to more detailed 

 analysis of these statistics, we surveyed some of 

 the main features of the data, of which some are 

 tabulated in Table 2. 



First we considered the direction of movement. 

 Figure 2 contains a synthesis of these data and 

 shows the direction of movement by tagging loca- 

 tion and time at liberty. It is implicit that we 

 treated each symbol as reflecting the behavior of a 

 sample of fish from the same statistical popula- 

 tion. The main features of Figure 2 are that fish 



tagged off New Jersey in July tended to move in an 

 eastward direction and both north and south dur- 

 ing the first 2 wk at liberty, but then movement 

 became strongly directed to the northeast. Fish 

 released in the Long Island area initially tended to 

 move toward the north, but after the first 30 days, 

 their movement appeared to have been concen- 

 trated to the west and south. The same conclusion 

 may be obtained from the southern New England 

 releases. 



An examination of the mean distance (Figure 2) 

 shows fish released off New Jersey tended to be 

 recovered at a slightly greater distance to the 

 north than the south. The fish moved approxi- 

 mately 7 or 8 miles per day. By the second 15-day 

 period, the fish moved about 60-100 miles to the 

 northeast. The pattern for Long Island releases, 

 based on only a few observations, shows that 

 movement distance of these fish during 1-15 days 

 was approximately the same as that for the New 

 Jersey releases. The short-term recoveries of 

 southern New England tagged fish refiect even 

 less average distances than New Jersey short- 

 term releases suggesting that either the fish off 

 southern New England moved less than off New 



901 



