ACOUSTIC TELEMETRY FROM FISH 



John Kanwisher,' Kenneth Lawson,' and Gunnar Sundnes^ 



ABSTRACT 



Methods are described for monitoring physiological parameters such as temperature and electrocardio- 

 gram from free swimming fish. Information is telemetered as sound radiating from an acoustic transmitter 

 implanted on the fish. Limitations of the technique and construction details of representative devices are 

 covered. Uses in both behavior and physiology are considered. 



Acoustic telemetry allows an investigator to 

 study the behavior and physiology of fish under 

 conditions which approximate their natural 

 state. Improvements in electronic techniques 

 permit construction of devices the size of one's 

 little finger; these devices can transmit data such 

 as heartbeat and temperature over ranges of sev- 

 eral hundred meters for as long as a month. We 

 describe here the use and constraints on sound as 

 a means of transmitting these data. We then dis- 

 cuss, in detail sufficient for duplication, the con- 

 struction of sample devices for transmitting, re- 

 ceiving, and interpreting the data. Finally, we 

 show how these devices have been applied to 

 specific experimental problems, and discuss the 

 results we have obtained. 



SOUND AS A TELEMETRY MEDIUM 



For ranges beyond a few meters through water, 

 sound is the only practical form of energy for 

 telemetry. It travels with little loss, whereas 

 radio waves and light are rapidly absorbed. Sev- 

 eral properties of sound in water are important. 

 For example greater ranges are possible in fresh 

 water than salt (one rarely has a choice in this). 

 Low frequencies transmit further than high. For 

 ranges up to several hundred meters, any fre- 

 quency below 100 kHz is suitable. If a range of 

 several kilometers is needed, the frequency 

 should be less than 20 kHz. Low frequencies, 

 however, involve longer wave lengths which im- 

 plies larger transducers. In the small devices 



necessary for fish work these are difficult to use. 

 Thus we most frequently employ frequencies be- 

 tween 40 and 80 kHz. Only in large tuna could we 

 use a transmitter big enough to work efficiently 

 at 20 kHz. It had an open sea range of 8 km. 



The interfering background noise, which 

 tends to obscure the signal, varies greatly at dif- 

 ferent places. In general, the shallow water 

 tropics are noisiest. At Coconut Island in Hawaii 

 the natural acoustic energy may be 100 times 

 greater than that at Friday Harbor in Puget 

 Sound. Most of the noise appears to be from bot- 

 tom animals such as snapping shrimp. Man-made 

 noise, like that from boat motors, can also be 

 troublesome. 



Relative motion between a sound source and 

 the receiver produces a Doppler shift in the ap- 

 parent frequency such that 



^/" relative velocity 



'Woods Hole Oceanographic Institution, Woods Hole, MA 

 02543. Contribution No. 3277 from the Woods Hole Oceanographic 

 Institution. This work was supported by National Science Founda- 

 tion Grant G A 31987X. 



^University of Trondheim, Trondheim, Norway. 



/ velocity of sound in water 



The velocity of sound in water is 1,500 m/s. A 

 relative velocity of 1 knot shifts frequency 0.03%. 

 This is only significant when frequency is inter- 

 preted critically, as in the depth transmitter to be 

 described. 



Additional complications arise from the inter- 

 ference effects due to multiple sound paths be- 

 tween transmitter and receiver. These are fre- 

 quently troublesome in small enclosures where 

 sound reflects from the walls. Nulls in the sound 

 field are produced which represent momentary 

 loss of signal. The ear has little trouble interpret- 

 ing periodic signals such as electrocardiogram 

 (EKG), but in a transcribed record these effects 

 can be confusing (see Figure 2). 



These remarks are meant to make one's ambi- 

 tions more modest when considering acoustic 



Manuscript accept October 1973. 



FISHERY BULLETIN: VOL. 72, NO. 2, 1974. 



251 



