FISHERY BULLETIN: VOL. 69, NO. 3 



from Station A and the reaction of the whales 

 during the 10th trial made it necessary to use 

 a second transducer simultaneously at Station B. 

 A copy of the Station A playback tape was used 

 at Station B, but it was not possible to synchro- 

 nize the playback sounds from the two projec- 

 tors. In the subsequent four trials, we turned 

 on the playback as soon as the tide changed, 

 thus eliminating the possibility of the whales 

 going by us unnoticed before the start of trans- 

 mission. The whales stayed at least 1.5 km down 

 the river from the sound sources during the four 

 transmissions using both projectors. 



In each of the seven control trials, when no 

 sound was played, the whales moved directly 

 up the river with the incoming tide, past the 

 transmitting stations. They behaved the same 

 as they did before any of our experiments, with 

 many individuals passing near the pier at Station 

 A. However, we could see very few whales from 

 the viewing station during the killer whale play- 

 baclis. Low-altitude reconnaissance flights con- 

 firmed our suspicion that numerous whales were 

 remaining well down the river out of view dur- 

 ing the transmissions. 



Belugas vocalize extensively (Schevill and 

 Lawrence, 1949; Fish and Mowbray, 1962). 

 We recorded their sounds many times at various 

 places throughout the river, often under ideal 

 conditions of flat-calm water with no boat traflic. 

 Once, we drifted in a small boat for over 2 hr 

 during a no-playback control with a group of 

 about 500 whales and recorded a spectacular va- 

 riety of vocalizations from this relatively un- 

 disturbed herd. The belugas emitted very few 

 sounds, however, when the killer whale signals 

 were being transmitted. Quieting of belugas 

 in the presence of killer whales was noted bj' 

 Schevill (1964). 



CONCLUSIONS AND 

 RECOMMENDATIONS 



Our experiments showed that playback sounds 



of killer whales can be used to keep belugas out 



of Alaska's Kvichak River. This method was 



very effective and practical. Installing and 



^maintaining such a playback system for 2 to 3 



weeks each year would seem to be an economi- 

 cally feasible way of reducing beluga predation 

 on red salmon smolts. Such a system could be 

 started automatically at the beginning of the 

 flood tides and, if necessary, left on for their 

 duration, both day and night. It would not be 

 seriously affected by adverse weather conditions, 

 except for a possible reduction in the range over 

 which the belugas could hear the sounds — be- 

 cause of higher ambient noise levels in the river 

 from wind and rough water. 



We recommend using two sound projectors, 

 properly situated to provide good signal levels 

 all the way across the river. Initially, our trials 

 with one projector stopped the whales, but the 

 belugas eventually learned to go up the opposite 

 side of the river where the playback signals 

 were very weak or nonexistent. There were 

 too many sand bars in the river to achieve good 

 signal levels across the river with one projector, 

 regardless of its source level. 



Although we only used the two-projector sys- 

 tem for four trials, we do not feel the whales 

 would habituate and ignore the playback sounds 

 with continued use. The avoidance of the belu- 

 gas to this system was striking. The fact that 

 the whales began swimming up the river again 

 after nine trials using the single projector was 

 due, we feel, to their learning to avoid the 

 source, rather than to habituating to the sounds. 

 With sound projectors located on both sides of 

 the river there is probably no channel where 

 the whales can go up the river without hearing 

 the playback sounds. On the other hand, even if 

 the belugas were to habituate after 2 weeks of 

 playback, this would not significantly affect the 

 usefulness of the technique. Playback could be 

 timed to coincide with the 2-week peak of the 

 smolt run. 



After completing this experiment with killer 

 whale playbacks, we projected a 2500-Hz con- 

 tinuous tone and 2500-Hz randomly pulsed tones. 

 The belugas continued up the river during the 

 continuous-tone playback, but turned back on 

 the two occasions when pulsed tones were trans- 

 mitted. Since these playbacks were tried after 

 the white whales had been subjected to the kill- 

 er whale sounds for 2 weeks, we cannot spec- 

 ulate on how naive whales would have reacted. 



534 



