Abstract
Transitioning to renewable energy is essential to mitigate climate change and improve energy security; however, it is critical to understand potential impacts on wildlife. There is evidence of spatial redistribution in marine mammals during the high intensity sounds of pile driving construction activity, but the behavioural reactions of individuals is currently unknown, limiting the ability to predict energetic consequences, and robustly quantify or mitigate population-level impacts. Here, we address this knowledge gap using GPS tracking data (24 individuals) at the main centre of abundance for harbour seals (Phoca vitulina) in England; a Special Area of Conservation (SAC). We used Mahalanobis distance to identify instances of unusual movements, and quantified the association with estimates of received sound levels from piling. In total, there were 216 encounters between seals and elevated sound levels associated with piling, with 15 unusual movement instances detected. These consisted of either (1) high speeds, (2) cessation of horizontal movement or (3) suddenly initiating movement. The estimated mean population-level response threshold was 186 (95% CI: 169–199) dB re 1 μPa2·s. This study provides vital information which will facilitate environmentally sustainable development by (1) providing a dose–response relationship for environmental impact assessments to estimate the probability of responses, (2) enabling characterisation of responses for simulation-based assessments and evaluation of mitigation strategies, and (3) facilitating more robust links between piling sound levels, individual behavioural reactions, energetic consequences, and ultimately population-level impacts. Understanding how responses may vary across geographic contexts (e.g., sites with less constricted haulouts) and for populations with less prior anthropogenic exposure are critical areas of future research. More broadly, our study improves understanding of marine mammal responses to underwater sound, and our approach applying Mahalanobis distance to pinniped tracking data demonstrates a general framework which can be used to identify unusual movements across different disturbance sources and taxa.