and not a result of a combination of two sounds having only one amplitude mode 

 each but at different frequencies. 



An indication of the origin of the background noise came to light during a trip 

 to Quintana Roo in August 1970. A student, Chris Tresslar, and I were snorkl- 

 ing on a reef off Puerto Morelos. Chris, a geologist, was banging on the reef 

 with his pick. He noticed that occasionally during this pounding a flush of 

 crackling became apparent. Upon investigation, we found that there are certain 

 places on the reef which, when pounded upon, would emit sounds similar to the 

 frying pan noises discussed above. Apparently, this is a result of disturbance 

 of localized pop' ^.ation of small sound producers, possibly snapping shrimp. 



Purposeful Sounds 



Staccatos and coos: 



These two apparent sounds are lumped here because it is evident that they are in 

 fact the same. The staccato represented in Figure 3B and 4A was taken from a 

 sequence of sounds produced spontaneously by a single squirrelf ish, probably 

 H. ascensionis , as it approached and passed the hydrophone. During this sequence 

 the cooing sounds were quite evident. When both were represented on a sonogram, 

 they were found to possess certain features in common, confirming a previous 

 opinion that the cooing sounds are simply staccatos some distance away from the 

 hydrophone. 



The staccato is composed in this case of 10 pulses. The first two pulses are 

 run together but there is a gradual widening of the time interval between pulses 

 and the centers of the final two are separated by a .06 sec. time interval. 

 When the recording was analyzed after slowing the tape to 1/4 speed (from 7 1/2 

 inches per second to 1 7/8 inches per second) each of the 10 individual pulses 

 could be further resolved into 5 to 7 apparent subpulses with their centers 

 separated by approximately .0025 sec. time intervals. 



Temporally this provides us with a picture of a rather complex superposition of 

 pulse trains. The basic subpulses seem quite regularly spaced temporally, indi- 

 cating that possibly the time interval between subpulses is governed by some 

 mechanical aspect concerning operation of the sound producing mechanism, viz., 

 swim bladder and associated muscles and bones. Variation in number of subpulses 

 per pulse may indicate a control mechanism employed by the fish to vary the 

 actual duration of each pulse of the staccato sound. 



The number of pulses per staccato sound, although 10 in the case of the sound 

 represented in Figure 3B and Figure 4A, is quite variable within the same indi- 

 vidual. During the sequence in question, the fish produced a staccato consist- 

 ing of only eight pulses which was followed 1.7 sec. later by a sound composed 

 of seven pulses. It appeared that the staccato sounds consisting of the greatest 

 number of pulses were separated from one another by the greatest time interval 

 but this is not definitely confirmed. 



It is probably safe to assume that the Longjaw Squirrelfish, H. ascensionis , 

 while spontaneously producing staccato sounds at night, is capable of modulating 

 the frequency of occurrence of the sounds, the number of pulses, and possibly 

 the number of subpulses and, therefore, the length of pulses. 



VI-90 



