72 



Fishery Bulletin 98(1) 



a. 



a 



Station number 



36 37 38 40 42 44 46 



200 



Figure 7 



Section of isotherms along a transect occupied on 3 June 1997, Bar above 

 station numbers indicates the Gulf Stream frontal zone and the location of 

 larval swordfish collections. 



m/s (Olson et al.. 1994), larvae that were 4 to 5 mm 

 SL in the Atlantic could have been transported from 

 as far away as 900 km. A similar trajectory was pro- 

 jected for small larvae of bluefin tuna, Tim units tliyn- 

 nus (McGowan and Richards, 1989). Larvae that were 

 <5 mm in length, collected off North Carolina could 

 have been spawned in the Florida Straits if they re- 

 mained in the core of the Gulf Stream. Current ve- 

 locities within the western Gulf Stream frontal zone, 

 where larvae most frequently reside, are less than 

 axial velocities (Lillibridge et al., 1990; Song et al., 

 1995; Limouzy-Paris et al., 1997). Further, departures 

 from a smooth, along slope. Gulf Stream trajectory, 

 in the form of meanders, intrusions, and filaments 

 along the western Gulf Stream frontal zone are fre- 

 quent (Pietrafesa, 1989). We collected swordfish lar- 

 vae frequently within these Gulf Stream anomalies 



(Figs. 4-7). Water within these features veers and re- 

 verses direction (Lee et al., 1991), the result being 

 that the northward translocation of swordfish larvae 

 within the frontal zone is checked and their north- 

 ward transport delayed. The possibility of spawning 

 between Cape Canaveral and Cape Hatteras is real, 

 but not certain. 



Acknowledgments 



We thank J. A. Hare and L. R. Settle (NOAA) for the 

 collection of some of the swordfish specimens reported 

 herein, J. A. Hare for manipulation of satellite im- 

 ages, E. H. Laban for the extraction of otoliths, and 

 C. W. Lewis, R. L. Ferguson, and J. D. Christensen for 

 preparation of figures. J. A. Hare, J. V. Merriner, and 



