exceed 200 cm fork length (about 120 kg), and are, 

 therefore, less likely to be found in cold water. How- 

 ever, the males may tend to remain in even warmer 

 water as they predominate (67-100%) in catches 

 farther south, particularly in the Caribbean and ad- 

 jacent regions. 



The variation in distribution by size in the north- 

 ern regions is apparently due to differences in feed- 

 ing habits coupled with temperature tolerances. 

 Swordfish over deep water feed largely on surface 

 animals (flying fish, etc.), local near-surface school- 

 ing species (herring, mackerel, etc.), mid-water, but 

 usually vertically migrating species (lanternfish, 

 barracudinas, etc.), and upon squids. In shallower 

 water, large swordfish, whilst also taking near- 

 surface species, make feeding excursions to the bot- 

 tom where the temperature may be as low as 5-10°C, 

 and feed upon redfish, hake, butterfish, and other 

 benthic species. These fish then apparently return to 

 the upper mixed layer while digesting their meal, 

 presumably to obtain a higher body temperature, 

 since there is no evidence of homoiothermy, or ele- 

 vated values, in this species. It is at this time that fish 

 may be seen near the surface on calm sunny days, 

 conditions that result in water temperatures that are 

 higher right at the surface. Swordfish harpooned at 

 the surface either have full stomachs or empty ones. 

 These latter are completely empty without even the 

 normal complement of nematodes or fish and squid 

 hard parts, a fact suggesting voiding of the contents 

 while the fish struggled against the harpoon line. 

 Swordfish have been observed from submersibles, 

 at depths of 500 m or more, and even to have been 

 apparently resting at, or near, the bottom. It is im- 

 possible to determine whether these fish were on 

 temporary excursions into these depths and low 

 temperatures, or whether they regularly remain in 

 this environment. 



SPAWNING 



The reproductive cycle of swordfish in the north- 

 west Atlantic appears to involve spawning to the 

 south, in the Caribbean and adjacent areas, where 

 the temperature exceeds 24°C. The vast majority of 

 gonads from fish captured north of lat. 35°N (Cape 

 Hatteras) have been in the quiescent stage, with ova 

 diameters less than 0.18 mm. Maturing ova may 

 exceed 1.0 mm. Occasional fish have been reported 

 with ripening ovaries (Fish, 1926; FRB unpublished) 

 but these are rare, numbering one or two a year, at 

 most. Similarly some milt has been noted in a few 



males, but this is not necessarily a sign of imminent 

 spawning. Fish (1926) estimated that a mature 

 female contained 16 million eggs, while another 

 specimen v.'as calculated to contain 5 million. 



SIZE 



The largest swordfish, the size of which can be 

 verified, was a fish of 915 lb. dressed weight (ap- 

 proximately 550 kg live weight) landed in Cape Bre- 

 ton. The average weight taken by the commercial 

 fishery, however, was much less than this, being 

 close to 120 kg (round) for harpooned fish, and in 

 1970, as low as 54 kg for all fishing methods. The 

 average size had fallen considerably since the intro- 

 duction of longlining in 1962 (Tibbo and Sreedha- 

 ran, 1974). The size distribution of commercial land- 

 ings during 1970 (Fig. 2) shows a peak frequency in 



Commercial Swordfish Landings 



1970 



, (N = I4089) 



963 

 (N=7732) 



200 300 



Drosed Weight ( Pounds! 



' ' I ' 

 400 



"" t T ( ' I 



1-" 

 900 



Figure 2. — Size distribution of swordfish landed in Canada 

 in 1970. (Dressed weight to live weight conversion factor 

 1.326.) 



the 41-50 lb (18.6-22.8 kg) dressed weight class. This 

 is equivalent (x 1.326) to 55-66 lb (24.9-30.0 kg) 

 round weight. 



SIZE/WEIGHT AND GROWTH 



Analysis of the relationship between fork length 

 (cm) and live weight (lb) ratio by the least squares 

 method, indicates slope coefficients of 2.6-3.1 for 

 different samples at different seasons, with correla- 

 tion coefficients higher than 0.9 



The rate of growth has been investigated in a 

 number of ways but no firm figures are available. 

 There are no scales in adults, the otoliths are minute, 



104 



