64 



Abstract.— The retrospective assign- 

 ment of collections of larval swordfish, 

 Xiphias gladiiis, taken from 1973 to 

 1980, to water tj^pes and area of the Gulf 

 Stream front, as well as three sets of 

 contemporary collections taken in 1984, 

 1988, and 1997, indicated that larvae 

 were collected most frequently within 

 the western Gulf Stream frontal zone. 

 Larval swordfish accumulate by local- 

 ized hydrodynamic convergence, rather 

 than localized spawning, and thus these 

 rare, surface-oriented larvae are found 

 more frequently within the frontal zone. 

 Lengths of larval swordfish taken from 

 curatorial collections, from contemporary 

 collections, and from published records 

 from the Caribbean Sea, the Gulf of 

 Mexico, and the western North Atlan- 

 tic, as well as assumptions about growth 

 rates and Gulf Stream transport, indi- 

 cated that swordfish may spawn as far 

 north as Cape Hatteras. 



Distribution of larval swordfish, 

 Xiphias gladius, and probable spawning 

 ofjp the southeastern United States 



John Jeffrey Govoni 



National Marine Fisheries Service, NOAA 



Southeast Fisheries Science Center 



Beaufort Laboratory 



101 Pivers Island Road 



Beaufort, North Carolina 28516-9722 



E-mail address Jeff Govonifflnoaa gov 



Bruce W. Stender 



Oleg Pashuk 



Marine Resources Research Institute 



South Carolina Wildlife and Manne Resources Department 



Charleston, South Carolina 29412 



Manuscript accepted 22 June 1999. 

 Fish. Bull. 98:64-74(2000). 



Swordfish, Xfp/z/os^/arf/f/s, is a cos- 

 mopolitan and highly migratory spe- 

 cies that spawns year-round (Grail 

 et al., 1983). Larvae of this fish do 

 not account for high numbers of the 

 ichthyoplankton or ichthyoneuston, 

 and as a consequence, data that de- 

 scribe their spatial distribution are 

 sparse. Grail et al. ( 1983) concluded 

 from available data that, in the west- 

 ern North Atlantic, small larvae (< 10 

 mm) occur most fi-equently in the 

 eastern Caribbean and in the Straits 

 of Yucatan and Florida in Novem- 

 ber, and that larger larvae (>10 mm) 

 occur most often there, as well as 

 in the Gulf Stream north to Cape 

 Hatteras from January to March. 

 Both length groups occur primarily 

 in surface water over depths deeper 

 than 200 m (Markle, 1974). Tibbo 

 and Lauzier (1969) first speculated 

 that larvae may be associated with 

 horizontal temperature and salinity 

 gradients. The collection of larvae 

 along the apparent edges of the Gulf 

 Loop Current in the Gulf of Mexico 

 (Richards and Potthoff, 1980) and 

 the Gulf Stream in the Atlantic (Pot- 

 thoff and Kelley, 1982; Post et al., 

 1997 ) supports the notion that larvae 



occur in greater abundance within 

 fi-ontal zones. Yet, the rarity of larvae 

 has hindered a clear understanding 

 of such coarse and fine scale (Haury 

 et al., 1978) spatial distribution. 



The scarcity of larval swordfish 

 has obscured an understanding of 

 their spawning pattern as well. The 

 mesoscale pattern of larval distribu- 

 tion (Grail et al, 1983) implies that 

 swordfish spawn in the Caribbean 

 and the Straits of Yucatan and Flor- 

 ida, and that their larvae are carried 

 northward by the Gulf Stream. Occa- 

 sional small larvae taken in the Atlan- 

 tic imply that swordfish may spawn as 

 far north as Cape Hatteras (Markle, 

 1974). Large, and presumably older 

 larvae in any location may be the re- 

 sult either of local spawning and sub- 

 sequent retention or of transport ft"om 

 a distant spawning locale. Small lar- 

 vae at a specific locale must be the ex- 

 clusive result of local spawning. 



Hydrated oocytes within the ova- 

 ries of adult females indicate that 

 swordfish spawn south of the Sar- 

 gasso Sea, in the northern Carib- 

 bean Sea, and the Straits of Florida 

 (Arocha and Lee, 1995), although 

 Squires ( 1962 ) has suggested that 



