FISHERY BULLETIN: VOL. 81, NO. 3 



groupings: 1) High concentrations east of the 

 Bahamas containing most leptocephali, 2) lower con- 

 centrations flanking this patch, and 3) Gulf Stream 

 larvae. 



Leptocephali started north and northwest of F (runs 

 A through D) showed more uniform distributions 

 than the previous month, although the sharp eastern 

 and southern gradients were maintained. There was 

 transport between the Bahamas and Cuba only in run 

 D, and leptocephali in run B showed lesser dispersal 

 when compared with other runs. Run C again showed 

 most clearly the effects of Antilles Current transport, 

 while run A showed a similar but less developed pat- 

 tern. Leptocephali in runs A and C reached lat. 39°N 

 and long. 65°W at concentrations >10" 7 . 



Run E continued to show the most widespread dis- 

 tribution, and while offshore of Florida there were 

 still concentrations > 10" 2 , the 10" 3 contour enclosed 

 a considerable area offshore of the mid-Atlantic 

 coast. 



Distribution of Leptocephali After 

 90 Days of Drift 



Runs A, C, D, and F became similar by the third 

 month of drift (Fig. 5; 30 May). The cluster analyses 

 for these simulations at day 90 are of interest, 

 because in each a distinct geographic grouping of 

 cells emerged consistently. These clusters did not 

 combine with others until forced to do so at the final 

 stages of clustering. This indicated that the cell 

 groupings represented by the clusters had spatial 

 distributions of leptocephali (as measured by 

 Moran' s I) which were internally more uniform than if 

 cells external to the groupings had been included 

 during clustering. This characterized (independently 

 of concentration contour plots) an inportant feature 

 in the spatial structure of the leptocephali at day 90. 

 The clusters of cells in runs A, C, D, and F formed 

 patches east of Florida and the Gulf Stream, north to 

 northwest of the Bahamas, and northwest of the 

 starting points (roughly between lat. 24° to 28°N and 

 long. 71° to77°W). These patches had mean cell con- 

 centrations of 0.015 to 0.025, and patch limits were 

 approximated by the 10 -2 contours (Fig. 5). The bulk 

 of the starting concentrations was contained within 

 these patches. 



Another feature common to the A, D, and F runs was 

 the large area of concentrations between 10 ;! and 

 10~ 4 that paralleled the mid- Atlantic coast and then 

 extended offshore at about lat. 38°N. Run C had an 

 identical pattern, except that its concentrations in 

 this area were an order of magnitude higher. This dif- 

 ference can be attributed to starting point C's loca- 



tion and the enhanced Antilles Current-Gulf Stream 

 transport mentioned earlier. 



Run B's pattern was similar to that of runs A, C, D, 

 and F, but was not as fully developed. This is because 

 run B leptocephali were started farther into the 

 Sargasso Sea, where currents are weaker. The main 

 patch of concentration was present, but it covered a 

 broader area and had a slightly lower mean concen- 

 tration of 0.01. The 10" 5 contour along the North 

 American coast in run B took the place of the 10~ 4 

 contour in runs A, D, and F. In B there were still high 

 concentrations near the starting point. 



Run E continued to exhibit the most extensive pat- 

 tern of leptocephalus distribution (Fig. 5E). There 

 was a patch offshore of Florida and the Gulf Stream 

 with a mean cell concentration of 0.012, but it was 

 smaller and farther north than the corresponding 

 patch in other runs. More than half of the starting 

 concentration lay north of lat. 31°N and within the 

 10~ 3 contour. Concentrations >10~ 2 were well north 

 of lat 32°N in run E. 



Distribution of Leptocephali After 

 150 Days of Drift 



The patterns of distribution below lat 32°N persisted 

 during the next several months in the simulations, 

 while to the north there was an increase in the con- 

 centrations and continued Gulf Stream transport. 

 The main patches of concentration offshore of the 

 Gulf Stream and Florida remained in runs A, C, D, 

 and F after 5 mo of drift (Fig. 6; 29 July). The patches 

 moved slightly to the northwest, and lay between lat. 

 25° to 30°N and long. 70° to 78°W. This region still 

 contained more than half of the total starting concen- 

 tration in runs A, C, D, and F, and the mean cell con- 

 centration in this area was about 0.005. 



The broad bands of concentration along the mid- 

 Atlantic coast also persisted and increased in con- 

 centration to between 10" 3 and 10 2 . This was a 

 consequence of the patches offshore of Florida just 

 mentioned. It appears the patches slowly introduced 

 larvae into the Gulf Stream system. The subsequent 

 transport in the Gulf Stream quickly spread larvae 

 parallel to the coast, then offshore and out into the 

 Atlantic. The patches in simulations A, C, D, and F 

 formed in the same general location, so that for larvae 

 carried north of lat. 30°N the starting location had a 

 lesser effect on the distribution. This is illustrated by 

 the fact that the 10~ 3 contours are virtually identical 

 in position north of lat. 30°N in runs A, D, and F at 5 

 mo after starting. Run C formed the patch earlier, so 

 more larvae had entered the Gulf Stream and were 

 therefore more widely dispersed. Leptocephali at 



490 



