HARBOUR PORPOISE RESPONSES TO PILE DRIVING ARE BETTER PREDICTED BY DISTANCE TO SOURCE THAN BY ENERGY-BASED RECEIVED SOUND LEVELS

Paul M Thompson, Aude Benhemma-Le Gall, Barbara Cheney, Virginia Iorio-Merlo, Oihane Fernandez-Betelu, Isla M Graham

1. Regulatory assessments for offshore construction are required to avoid impacts on protected marine mammals through noise-related injury or disturbance. Criteria for injury risk are widely accepted, but the extent to which behavioural responses are related to noise levels or co-varying contextual factors such as proximity remains uncertain.

2. This study used arrays of echolocation detectors to compare behavioural responses of harbour porpoises along gradients of exposure to pile installation noise at three offshore wind farms. Differences in the size of piles and installation hammers at these sites meant that received levels at given distances differed by up to 15 dB, providing an opportunity to assess whether behavioural responses to pile installation noise were best predicted by received doses of impulsive noise or distance to source.

3. Using balanced data from within 35 km of the first two piling bouts at each site, the probability of detecting a response was better explained by the distance to the piling vessel than either weighted or unweighted sound exposure levels of impulsive piling noise. Response functions for individual wind farms were also more similar for distance functions (50 % probability of response at 6.5, 4 & 5.5 km) compared to those unweighted single strike Sound Exposure Levels (50 % probability of response at 148.5, 152 & 166.5 dB re 1 µPa2s). This suggests that energy-based metrics do not capture all the acoustic cues that best explain behavioural responses to pile driving noise.

4. Synthesis & applications. Regulatory assessments for increasingly large offshore wind farms have been constrained by uncertainty over the most appropriate behavioural response functions for estimating disturbance. Earlier single site studies found similar support for response functions based on both received noise levels and distance. Some UK Statutory Nature Conservation Bodies have therefore recommended using a dose-response as a precautionary measure, but extrapolation has subsequently led to unrealistic predictions of far-field disturbance. There is growing evidence that marine mammal responses to other noise sources are highly dependent upon proximity to source. Our findings suggest that distance-based response functions provide a more consistent and practicable variable for assessing disturbance for a range of different foundation types and source levels.