A series of preliminary tests were run in order to set the volume of
the projected stimuli. Several combinations were tested using the
same experimental settings in terms of distance and position of the
speaker to the recording hydrophone (Reson TC4032), the latter
‘mimicking’ the position of the tested fish. The played-back noise was
re-recorded, the power spectra calculated with Spectra RTA (Sound
Technology) spectral analyser and the spectra superimposed on the spectra of the boat noises, in order to compare SPLs and frequency
distributions; the stimuli with the most similar spectrum to the
recorded one were chosen. Finally, the particle acceleration level of
the projected noises at each frequency was calculated applying
spherical-field Eq. (8).