Evaluation of a Low-Frequency Sound-Pressure System for Guiding Juvenile Salmon away from Turbines at Bonneville Dam, Columbia River

Abstract

In June 1995, we evaluated the effectiveness of a 122-m-long array of 25 low-frequency transducers for guiding juvenile salmon away from turbine units 9 and 10 at Powerhouse I of Bonneville Dam, Columbia River, Oregon. Juvenile salmonids included subyearling and yearling chinook salmon Oncorhynchus tshawytscha and yearling coho salmon O. kisutch, steelhead O. mykiss, and sockeye salmon O. nerka. Generated sounds were predominated by 300-Hz and 400-Hz frequencies and transmitted as 2-s crescendos, with repeated amplitude ramps from 0 to about 160 decibels referenced to 1 μPa at 1 m every 2 s. Sound-on and sound-off treatments did not differ significantly in the mean number of fish passing north or south across the upstream end of the array, where the angle of incidence of flow was only about 5 degrees. The power of these one-tailed t-tests (α = 0.05) for detecting 50% differences in means was 82% for fish passing north across the array and 99% for fish passing to the south. We also counted smolts in front of four turbine intakes using fixed-aspect hydroacoustic equipment and found no significant differences in the mean number of smolts upstream of intakes during 4-h sound-on and sound-off treatments. The statistical power of 4-h tests was 98% or greater for detecting differences in means as small as 20% (α = 0.05 in a two-tailed analysis of variance and a one-tailed t-test). To help corroborate and interpret the negative results obtained in the field experiment, in 1997 we evaluated reactions of captive schools of subyearling chinook and coho salmon and yearling sockeye in a net-pen to the same 300–400-Hz signal. We observed no startle reactions and found that the frequency of avoidance of the signal was no greater than the frequency of coincidental avoidance during control trials without sound. After exposure to the 300–400-Hz signal, one school of subyearling chinook exhibited nondirectional startle responses to 150-Hz or 180-Hz sound, indicating that those fish could respond. We conclude that the 300–400-Hz signal did not influence the behavior or distribution of juvenile salmon in either study.