CAREY and ROBISON: DAILY PATTERNS IN ACTIVITIES OF SWORDFISH 



Figure l. — Depth transmitter for swordfish. Rubber bands hold the transmitter to an adapter which fits on a standard swordfish 

 harpoon. The crossbar on the adapter serves to limit penetration of the harpoon dart. A conductive path between two elements of a 

 seawater switch turns the transmitter on when it enters the water The coin is 24 mm in diameter 



Changing battery voltage did not affect the pulse 

 rate of either type of transmitter, and the pulse 

 rate of depth transmitters was not affected by 

 temperature changes in the range 5°-30° C. In use 

 we found that several days into an experiment the 

 depth transmitters would still indicate < 1 m when 

 the fish was seen on the surface. A temperature 

 transmitter recovered after 13 d on a fish was 

 recalibrated and found to be within 0.1° C of the 

 original calibration. 



Receiving 



We listened to the transmitters with a direc- 

 tional hydrophone constructed of seven ceramic 

 transducer rings wired in parallel in a cylindrical 

 array 10.5 cm long. The rings were backed with a 

 layer of closed-cell polyethylene foam and cast 

 in epoxy. A preamplifier in the hydrophone ampli- 

 fied the signal 100 x and reduced problems from 

 electrical noise picked up on the cables. The upper 

 and rear surfaces of the horizontally mounted 

 hydrophone cylinder were acoustically shielded 

 with four alternating layers of 3 mm thick foam 

 rubber and steel, leaving an approximately 140° 

 sector uncovered and facing forward and down. 

 The hydrophone was mounted at the bow of the 

 vessel on a streamlined fin arranged so that it 

 could pivot and follow the flow of water as the boat 

 yawed. A shaft through the fin allowed the hydro- 

 phone to be rotated by a rope and pulley system 

 from the bridge of the vessel. 



The receivers (Lawson Instrument Company 

 and CAI CR-40) were mounted on the bridge so 



that one person could follow the fish by rotating 

 the hydrophone to find the strongest signal, then 

 steering the boat in that direction. Data were 

 recorded by timing a fixed number of pulses to the 

 nearest 0.1 s with a stopwatch and converting the 

 pulse rate to temperature or depth using a linear 

 regression for the calibration curve. A 0.1 s count- 

 ing error in a 30 s count produced a depth error of 

 approximately 2 m. Data were taken routinely 

 every 5 or 10 min and more frequently when the 

 fish was changing depth. An automatic data 

 recording system, based on a phase lock receiver, 

 was used on 11 November 1977 and gave a contin- 

 uous, detailed record of a swordfish rising to the 

 surface in a series of steps. 



Navigation 



In the Baja California area we used radar to 

 determine range and bearing to various peaks and 

 headlands. The relative accuracy of this technique 

 was good as checked by comparison of echo sound- 

 er depth with depth given on the chart at the 

 plotted position. In the experiment near Cape 

 Hatteras we used loran C with an accuracy better 

 than 1 km. The position of the vessel was recorded 

 when it was brought close to the fish and the plot 

 of these positions used to approximate the fish's 

 course. 



Temperature 



Temperature of the water was measured with 

 expendable bathythermograph probes (XBT) (Sip- 



279 



