406 MECHANICAL AND ACOUSTICAL SENSES 



sharks were adults or subadults. These responses indicated that the with- 

 drawal responses noted by Banner (1972) were not due to the shallow depth 

 at his test site or to his young subjects. 



Two years later we began to explore this phenomenon more fully with the 

 idea that a successful predator, when approaching a source transmitting 

 signals having appropriate acoustical properties, might well change its be- 

 havior when the properties to which it was attending suddenly changed in an 

 "unexpected" manner. Since predatory mechanisms, behavior or otherwise, 

 obviously require energy expenditure, an appropriate and highly adaptive 

 mechanism for any predator would be one that would aid "go-no go" 

 decisions regarding problems involving conservation of energy in such con- 

 texts. If such a mechanism were operating, sudden changes in the expected 

 flow of information being processed at the moment could rapidly influence 

 ongoing behavior. 



The results of the ensuing studies will appear elsewhere, but a brief sum- 

 mary follows. It became clear that rapid withdrawal could be elicited in 

 adult lemon and silky sharks by projecting underwater sounds similar to 

 those eliciting approach in the first place. The frequency spectrum was the 

 same in both types of sounds (i.e., broadband, covering one or more octaves 

 within the hearing range). Also, both types could be pulsed. There were dif- 

 ferences, however; sounds eliciting withdrawal possessed rather long intervals 

 (e.g., 2.5 s on, 2.5 s off). This distinction paralleled that noted by Banner 

 (1972). 



An additional factor, mentioned by Banner as perhaps playing a role in 

 the response, appeared to be most important. This concerned a sudden in- 

 crease in the sound level as a subject approached the given source. For 

 example, a sudden increase of 15 to 20 dB when a shark was within a few 

 meters of the source resulted in the animal's retreating at a speed often 

 faster than its initial approach. This response has been frequently documen- 

 ted from captive adult lemon sharks, held under conditions adequate for 

 appropriate testing (Klimley 1976), as well as by young and adult free- 

 ranging silky sharks found in the offshore waters of the Straits of Florida 

 and the Tongue of the Ocean, Bahamas (Myrberg et al. 1975b). The basic 

 response pattern was similar in both species; the restrictions imposed on the 

 movements of the lemon sharks by the test facility (a large channel) resulted, 

 however, in a more stereotyped form of withdrawal. Basically, this consisted 

 of the shark moving in a reasonably narrow arc and proceeding in the direc- 

 tion from which it came. The pattern shown by silky sharks was more vari- 

 able—probably due to the lack of any restrictions on their movements. The 

 general response was neither a simple "startle" nor a series of rapid turns or 

 sequences of apparently disorganized movements; rather, it consisted of a 

 sequence of movements resulting in the animal(s) disappearing from view 

 within 10 to 30 s of the change in the ongoing sound. 



The orientation of such movements also showed that the animals recog- 

 nized the direction of the sound source. Some individuals turned within a 

 second or two of signal change through a narrow arc and headed out of 

 visual range; others that had been heading directly at the transducer veered 



