FISHERY BULLETIN: VOL, 79, NO. 2 



Examination of swordfish stomach contents 

 (Scott and Tibbo 1968; Ovchinnikov 1970) showed 

 that they feed on cephalopods and a wide variety of 

 nektonic fishes (e.g., anchovies, hake, mackerel) 

 and micronektonic fishes (e.g., myctophids, para- 

 lepidids) and that a significant portion of their diet 

 consisted of vertically migratory mesopelagic spe- 

 cies. We have also found unusual items in their 

 stomachs including birds and shrimp. Swordfish 

 have been observed swimming through schools of 

 fish, stunning them with their swords before 

 eating them (Goode 1883). Fish removed from 

 swordfish stomachs often show damage to the 

 muscle and vertebral column from having been hit 

 by the sword (Scott and Tibbo 1968). Swordfish 

 have large eyes and are efficient visual predators 

 even in dim light. Small phosphorescent lights are 

 occasionally used to attract them to longline 

 hooks, which indicates that they may also respond 

 to bioluminescence. 



Swordfish are aggressive and there are many 

 accounts of their attacking and ramming their 

 bills into ships, whales, and other objects (Goode 

 1883; Gudger 1938; Smith 1956; Jonsgard 1962). 

 We have seen penetrating wounds in swordfish 

 and Edlin found a 15 cm fragment of a swordfish 

 bill that entered near the heart of a 70 kg 

 swordfish and was driven back into the body 

 cavity, which may indicate that they strike each 

 other. While they are generally solitary, longline 

 fishermen say that in the Straits of Florida, a 

 spawning area, swordfish may be encountered in 

 twos with some regularity. 



In the present study we investigate the horizon- 

 tal and vertical movements of swordfish during 

 their daily activities and attempt to determine the 

 range of temperatures they encounter. We used 

 acoustic telemetry to monitor data from depth and 

 temperature sensors attached to the fish for peri- 

 ods of up to 5 d. The results presented here provide 

 the first description of their activities based on 

 continuous direct observations of individuals. 



METHODS 



We attempted to attach transmitters to seven 

 swordfish and were successful with five in the 

 Pacific, near Cabo San Lucas at the tip of Baja 

 California, and one in the Atlantic, east of Cape 

 Hatteras, N.C. 



For the experiments off Baja California, the 

 swordfish were located on the surface by aircraft 

 and the transmitters harpooned into the free- 

 swimming fish from the tracking vessel Sea 

 World. In the Atlantic, the swordfish were taken 

 on commercial longline fishing gear set by the 

 tracking vessel FV Diane Marie. Weights of the 

 fish were estimated by the fishermen. 



Transmitters 



Two types of sensors were used in the trans- 

 mitters. Depth transmitters had a 500 or 1,000 

 Ib/in^g Biotek^ strain gage pressure transducer. 

 Temperature transmitters had a 300 kli Fenwall 

 GA53M2 thermistor linearized with a series resis- 

 tor. An up-down integrating circuit converted 

 resistance changes in the sensor to a varying pulse 

 rate. The pulses keyed an oscillator and output 

 stage which drove a 1.27 cm long, 2.79 cm (outer 

 diameter) cylindrical ceramic transducer (Marine 

 Research TCD 5) with 30 ms pulses of ultrasound 

 at an electrical power level of one to several watts. 

 The transducers, which were mechanically reso- 

 nant at 33 kHz, were operated at 32 kHz for 

 temperature and 34 kHz for depth so that signals 

 could be separated in experiments where both 

 were used simultaneously. 



Power was supplied by a battery of five 1.2-Ah 

 lithium cells (Mallory L0325) which give a useful 

 life of about 1 wk. Range was as great as 3-5 km at 

 times, but much shorter when propagation condi- 

 tions were poor. 



The transmitters were 14 cm long, 4.5 cm wide 

 by 3 cm thick and weighed 250 g in air and 90 g in 

 water. They were cast in a strong epoxy plastic 

 (Hysol 2039 resin, 3561 hardener) and tested to 

 withstand hydrostatic pressures equivalent to a 

 depth of 1,000 m. A miniature (6.5 cm long) 

 swordfish dart was tied to the end of the trans- 

 mitter with a 13 cm loop of twisted 200-lb (91 kg) 

 test monofilament nylon (Figure 1). An adapter 

 which fitted on a standard swordfish harpoon had 

 a crosspiece which limited penetration of the dart 

 to about 10 cm. In some experiments depth and 

 temperature transmitters were tied in tandem 

 and attached to the fish with a single dart. The 

 swordfish showed no obvious reaction to the tags 

 once they were attached. 



The instruments were stable and accurate. 



■'Robert Edlin, P.O. Box :341303, Coral Gables, FL 33134, pers. 

 commun. 1980. 



"^ Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



278 



