species were retained for determination of mercury 

 content. 



RESULTS 



Stomach analyses 



Sixteen families of fishes and the short-finned 

 squid (/. illecebrosus) were identified from 141 

 swordfish stomachs. One stomach contained the 

 remains of two octopi. 



Percentages by volume offish (all species) versus 

 squid in stomachs ranged from 78.7 to 94.0% (Table 

 1). These results are consistent with our earlier find- 

 ings of 68.4 to 86.2% (Scott and Tibbo, 1968) and 

 confirm the importance of fish in the diet of sword- 

 fish in the Northwest Atlantic. However, the 

 species of fishes and the amount of squid in stomachs 

 varied with feeding areas. In the Grand Bank and 

 Banquereau regions, twice as much squid (121.5 cc 

 average per stomach) occurred in stomachs as in 

 samples from Emerald Bank (62.9 cc average per 

 stomach) (Table 1). Also, in the Grand Bank region, 

 the volume of redfish eaten was greater than for any 

 other species, whereas the silver hake (M. hilinearis) 

 was absent from the diet. The sample from Emerald 

 Bank region, however, contained a greater volume 

 of silver hake than any other fish except the bait, 

 mackerel. 



Total and average volumes of all food in swordfish 

 stomachs for the different size groups are given in 

 Table 2. The figures for average volume within each 

 size group show that volumes increase with increase 

 in size of swordfish, as might be expected. The aver- 

 age volume of food within each size group was simi- 

 lar for both the 1964-65 and 1971 samples. 



In general swordfish feed on fewer fish species 

 and on more squid in the Grand Bank and Ban- 

 quereau regions than in those areas to the south and 

 west. The number offish species increases and the 



Table 1. — Volumes (cc) of fish and squid in swordfish 

 stomachs from 1971 samples. 



Table 2. — Average volumes of all food in swordfish 

 stomachs arranged by length groups of swordfish for 

 1964-65 and 1971 samples. 



amount of squid in the stomachs decreases in regions 

 to the south and west, particularly Browns and 

 Georges banks and offshore canyons such as 

 Lydonia, Hydrographer, and Washington. 



Fishes 



The 16 families of fishes eaten are listed in Table 3. 

 The first five food items are of primary importance 

 and constitute 84.7% by volume of the total fish 

 ingested. The remaining 1 1 families are of secondary 

 importance (6.9%) and, indeed, some of these, such 

 as the pearlsides (Maumlicus muelleri), may be rare 

 in the diet, since this is our first report of the species 

 from a swordfish stomach. Unidentified fishes con- 

 stituted 8.4% of the total. 



Mackerel (5. scombrus) deserves special mention 

 because it was used as bait. Also, there is evidence 

 of "bait robbing"; that is, two or more mackerel, 

 bearing evidence of hook marks, were found in a 

 single stomach, suggesting that swordfish success- 

 fully remove bait from hooks. The usual condition 

 was one mackerel, presumably bait, per stomach. 

 Occasionally, the remains of two mackerel were 

 present. On two occasions three occurred in a single 

 stomach and on one occasion five mackerel were 

 eaten by one swordfish. However, the state of diges- 

 tion often obscures hook marks. Special marking of 

 bait would be most helpful in determining the role of 

 mackerel in the natural diet of swordfish. Undoubt- 

 edly, the large volume of mackerel in the diet, rep- 

 resenting 36% of the total, is an unnatural condition. 



Barracudinas, family Paralepididae, were the 



139 



