Repeated exposure reduces the response to impulsive noise in European seabass

Abstract

Human activities have changed the acoustic environment of many terrestrial and aquatic ecosystems around the globe. Mounting evidence indicates that the resulting anthropogenic noise can impact the behaviour and physiology of at least some species in a range of taxa. However, the majority of experimental studies have considered only immediate responses to single, relatively short-term noise events. Repeated exposure to noise could lead to a heightened or lessened response. Here, we conduct two long-term (12 week), laboratory-based exposure experiments with European seabass (Dicentrarchus labrax) to examine how an initial impact of different sound types potentially changes over time. Naïve fish showed elevated ventilation rates, indicating heightened stress, in response to impulsive additional noise (playbacks of recordings of pile-driving and seismic surveys), but not to a more continuous additional noise source (playbacks of recordings of ship passes). However, fish exposed to playbacks of pile-driving or seismic noise for 12 weeks no longer responded with an elevated ventilation rate to the same noise type. Fish exposed long-term to playback of pile-driving noise also no longer responded to short-term playback of seismic noise. The lessened response after repeated exposure, likely driven by increased tolerance or a change in hearing threshold, helps explain why fish that experienced 12 weeks of impulsive noise showed no differences in stress, growth or mortality compared to those reared with exposure to ambient-noise playback. Considering how responses to anthropogenic noise change with repeated exposure is important both when assessing likely fitness consequences and the need for mitigation measures.

Keywords: Dicentrarchus labrax; European seabass; anthropogenic noise; growth; habituation; hearing threshold; pollution; stress; tolerance; ventilation rate.

Figure 1

(a–h) Illustrative spectrograms of the four sound types used in the experiments, showing both examples from an original recording and from the recording of playback in one of the long‐term exposure tanks. (i) Power spectral densities of sound pressure levels from recordings of original ambient and ship conditions and playback of those recordings in a long‐term exposure tank. Playbacks were affected by near‐field effects, and speaker performance meaning some frequencies were louder and others quieter, but ships were louder than ambient noise and ship‐noise playbacks were louder than ambient‐noise playbacks. Sounds <10 Hz are unlikely to be generated by the speakers, but may result from, for example, background pump noise or vibrations in the experimental laboratory. The higher levels at >1500 Hz for ambient‐noise playbacks compared to original ambient‐noise recordings likely result from background noise, the resonant frequency of the tank, and the frequency response of the playback set‐up.

Figure 2

Change in opercular beat rate of seabass in experimental set 1 during two consecutive short‐term (2 min) exposures to playback of recordings of different sounds (ambient noise, pile‐driving noise or ship noise). In (a) are responses of ‘naïve’ (no prior experience of playbacks) postlarval individuals to ambient‐noise playback followed by playback of one of the three sounds (n = 90 evenly spread between the three treatments). In (b–d) are responses of individuals that have experienced 12 weeks exposure to ambient‐noise playback, pile‐driving‐noise playback or ship‐noise playback, respectively; testing involved a change from playback of the long‐term noise exposure to a different playback track (n = 90 evenly spread between treatments in each case). Shown in all cases are means ± SE, with the significance of pairwise post hoc tests indicated above bars (significant results in bold).

Figure 3

Change in opercular beat rate of seabass in experimental set 2 during two consecutive short‐term (2 min) exposures to playback of recordings of different sounds (ambient noise, pile‐driving noise or seismic noise). In (a) are responses of ‘naïve’ (no prior experience of playbacks) postlarval individuals to ambient‐noise playback followed by playback of one of the three sounds (n = 90 evenly spread between the three treatments). In (b–d) are responses of individuals that have experienced 12 weeks exposure to ambient‐noise playback, pile‐driving‐noise playback or seismic‐noise playback, respectively; testing involved a change from playback of the long‐term noise exposure to a different playback track (n = 90 evenly spread between treatments in each case). Shown in all cases are means ± SE, with the significance of pairwise post hoc tests indicated above bars (significant results in bold).