Information on the distribution of ripe females is 

 available from both commercial and sportfishing 

 operations. The Georges Bank area has been heav- 

 ily fished, but there are very few reports of ripe 

 females from this region, although some females 

 bearing maturing eggs have been taken off the New 

 England coast (Fish, 1926; Lee, 1942; Rich, 1947). 

 In contrast, there are numerous accounts of ripe 

 females caught off the northern coast of Cuba 

 (Arata, 1954; Lamonte and Marcy, 1941). Accord- 

 ing to Lamonte (1944), fishermen and anglers claim 

 that swordfish bearing huge ovaries, with eggs 

 ready to rupture the ovigerous membranes, are fre- 

 quently found in the Cojimar, Cuba area, often ac- 

 companied by another much smaller fish, presuma- 

 bly the male. Such a distribution of ripe adult 

 swordfish suggests that spawning occurs some- 

 where off the north coast of Cuba, rather than much 

 farther north. The occurrence of small larvae in the 

 Florida Straits of Cape Hatteras region supports 

 this conclusion. However, a single spawning area 

 cannot account for the widespread distribution of 

 larvae. 



In the Western Atlantic, it is probable that 



swordfish spawn in widely scattered areas from 

 which the larvae are further dispersed by currents 

 such as the Gulf Stream. This contrasts with 

 Gorbunova's (1969) conclusion that swordfish 

 spawning in the Pacific is restricted to areas of up- 

 welling, where high productivity provides favorable 

 conditions for both zooplankton and fish feeders 

 such as larval swordfish. Gorbunova (1969) also 

 concluded that young swordfish do not migrate far 

 in the first year and thus are captured quite close to 

 the actual spawning grounds. 



In the Western Atlantic, Arata (1954) proposed a 

 large spawning area and an extended spawning 

 period. From larval sizes and the growth rate of 0.6 

 mm/day suggested by Sanzo (1922), Arata esti- 

 mated the approximate ages of his specimens, and 

 deduced that peak spawning occurs at approxi- 

 mately the same time in both the Gulf Stream and in 

 the Gulf of Mexico — from May to June in the Gulf 

 Stream and from late April to July in the Gulf of 

 Mexico. Arata (1954) also suggested that larvae 

 may be carried long distances in the Gulf Stream 

 system. Making back calculations based on fish 

 sizes and current speeds, and assuming passive drift 



Figure 2. — Numbers, size ranges, and mean lengths of swordfish larvae — Fisheries Re- 

 search Board of Canada collections. 



255 



