Efforts to meet climate change targets are resulting in rapid and global expansion of offshore windfarms. In many regions, development areas are also used by protected marine mammals, requiring the assessment and mitigation of any risk of injury during construction and operation. For small cetaceans such as the harbour porpoise, there is particular concern over the risk of injury from impulsive noise should individuals remain within near-field injury zones during the installation of pile driven turbine foundations. Currently, this risk is assessed by comparing predicted noise levels at the start of piling with baseline estimates of animal density, which are, in turn, based on data collected at least one year earlier. However, vessel-based preparation work immediately prior to piling may displace animals, thus reducing any risk of injury when pile-driving begins. We investigated the effects of pre-piling activities on local soundscapes and harbour porpoise occurrence during the construction of two deep-water offshore windfarms in NE Scotland. Arrays of echolocation click detectors deployed at a sub-set of turbine sites were used to assess porpoise occurrence within a 5 km buffer during a 48-h period prior to the initiation of piling. In parallel, we characterised local vessel activity using AIS data and underwater broadband noise levels. We then used daily engineering records to characterise variation in construction activities and explore how porpoise occurrence varied during the 48 hours prior to piling. On average, vessels arrived onsite 11-15 h before the start of pile-driving activities at both windfarms. In both installation campaigns, harbour porpoise occurrence gradually declined by up to 33% during the 48 h prior to piling. This decrease in detections was associated with increased levels of vessel and pre-piling installation activities, and increased local underwater broadband noise levels. These results provide strong evidence of porpoise displacement prior to active mitigation activities, highlighting the need to account for disturbance from multiple sources when optimising mitigation measures aimed at reducing impacts of windfarm construction on protected marine mammal populations.

Passive Acoustic Monitoring (PAM) data were collected using echolocation click detectors (CPODs) and underwater noise recorders (SoundTrap). These data were processed and the presence/absence of harbour porpoise clicks and mean Sound Pressure Levels were summarised by hour relative to piling (HRP), per deployment, site and in relation to piling events. Automated Identification System vessel-trcking data were purchased, processed and the vessel intensity, mean and minimum distance from vessels to PAM devices were summarised per HRP. Daily engineering records were provided by the windfarm developpers and used to calculate the proportions of activities conducted by the piling vessel per HRP and windfarm and estimate the time of arrival of the piling vessel at construction sites.

R can be used to open and run the R scripts and the R project. The .txt files can be opened with Notepad but can also be called directly in R using the R scripts. Funding provided by: University of AberdeenCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000882Award Number: Funding provided by: Marine Scotland Science*Crossref Funder Registry ID: Award Number: Funding provided by: Beatrice Offshore Windfarm Limited*Crossref Funder Registry ID: Award Number: Funding provided by: Moray Offshore Windfarm (East) Limited*Crossref Funder Registry ID: Award Number: