476 MECHANICAL AND ACOUSTICAL SENSES 



the usually very low recovery percentages. Also, the very variable and un- 

 predictable time lapse between tagging and recovery further reduces the 

 value of what little data are collected. A telemetering study using single- 

 radio-release units can be likened to a conventional tagging study in which 

 nearly 100% recovery occurs and in which recovery times can be precisely 

 chosen in advance by the tagger. Thus, blue sharks off Santa Catalina could 

 be tagged with single-radio units set to detach in 24 h (or 48 h, etc.), 

 and popup locations would be determined by receiving stations on island 

 mountaintops. Boats would then be directed to the site for transmitter 

 recovery, unless the location were so distant that cost of the recovery 

 mission would exceed the value of the transmitter. Longer term trackings 

 (1 or more weeks) would probably require searching by light airplane and, 

 furthermore, would be best accomplished by multiple-radio units. A package 

 of at least three radios would be feasible for blue sharks; even more could be 

 carried by larger species such as the great white shark. 



An exciting future possibility, should governmental support become avail- 

 able, is worldwide tracking from polar-orbiting satellites. An ambitious 

 project of this type was proposed by Goodman et al. (1973) for the satellite 

 tracking of migrating whales over periods of 1 year or more. The general 

 route of the animal during the year would be monitored by satellite-fixing 

 of periodically released expendable radio buoys (401 MHz). At the end of 

 the tracking a larger multichannel data-recording package would detach for 

 recovery by a waiting shipboard crew. 



Detection of Specific Behaviors 



Besides the telemetering of relatively general, long-term trends of movement, 

 activity, or depth, the researcher may also wish to identify the occurrence of 

 other more specific behaviors. Such "specific behavioral events" may occur 

 only briefly and infrequently and thus may be missed by the tracker unless 

 the sensor data are continuously recorded, either at the receiver end or in a 

 transmitter storage unit to be later interrogated. 



Feeding Behavior— Very little is known about how most large, 

 active sharks normally obtain their food. While some data exist on what 

 they eat (from gut contents, preferred baits), surprisingly little is known 

 about when they feed, how often they feed, and exactly how they capture 

 their normal prey. Telemetering techniques can be used to provide this kind 

 of information. 



One method would involve stomach-implanted transmitters that directly 

 detected the intake of food by sensing changes in pH or temperature when 

 food and accompanying water were swallowed, or when stomach secretions 

 increased. Another way would be by recognition of specific events such as 

 turns, accelerations, or jaw movements, which are probably correlated with 

 feeding actions. Multichannel telemetry increases the probability of iden- 

 tifying significant specific events because it allows close time comparisions 

 between two or more factors. For example, "sudden direction change" 

 (compass sensor) occurring during the same brief time period as "sudden 



