• Energy Research

1. Offshore windfarms require construction procedures that minimise impacts on protected marine mammals. Uncertainty over the efficacy of existing guidelines for mitigating near-field injury when pile-driving recently resulted in the development of alternative measures, which integrated the routine deployment of acoustic deterrent devices (ADD) into engineering installation procedures without prior monitoring by Marine Mammal Observers. 2. We conducted research around the installation of jacket foundations at the UK’s first deep-water offshore windfarm to address data gaps identified by regulators when consenting this new approach. Specifically, we aimed to a) measure the relationship between noise levels and hammer energy to inform assessments of near-field injury zones, b) assess the efficacy of ADDs to disperse harbour porpoises from these zones. 3. Distance from source had the biggest influence on received noise levels but, unexpectedly, received levels at any given distance were highest at low hammer energies. Modelling highlighted that this was because noise from pin pile installations was dominated by the strong negative relationship with pile penetration depth with only a weak positive relationship with hammer energy. 4. Acoustic detections of porpoises along a gradient of ADD exposure decreased in the 3-hours following a 15-minute ADD playback, with a 50% probability of response within 21.7 km. The minimum time to the first porpoise detection after playbacks was > 2 hours for sites within 1 km of the playback. 5. Our data suggest that the current regulatory focus on maximum hammer energies needs review, and future assessments of noise exposure should also consider foundation type. Despite higher piling noise levels than predicted, responses to ADD playback suggest mitigation was sufficiently conservative. Conversely, strong responses of porpoises to ADDs resulted in far-field disturbance beyond that required to mitigate injury. We recommend that risks to marine mammals can be further minimised by: 1) optimising ADD source signals and/or deployment schedules to minimise broad-scale disturbance; 2) minimising initial hammer energies when received noise levels were highest; 3) extending the initial phase of soft start with minimum hammer energies and low blow rates.Minhyuk Seo Data consist of the following 16 files/file packages. A full description of data identifiers, R code and the data files required to repeat each analysis is provided in the text file: Thompson_BOWL_cMMMP_Data_Packages_Readme.txt Raw data for analyses of echolocation detections are available at: https://doi.org/10.5061/dryad.5qg30sd 1_Thompson_BOWL_cMMMP_noise_wav_files_data_2020-10-12.zip 2_Thompson_BOWL_cMMMP_piling_pulse_data_2020-10-12.txt 3_Thompson _BOWL_cMMMP_R_code_to_analyse_piling_noise_variation_2020-10-12.R 4_Thompson_BOWL_cMMMP_piling_times_data_2020-10-12.txt 5_Thompson_BOWL_cMMMP_ADD_wav_files_data_2020-10-12.zip 6_Thompson_BOWL_cMMMP_R_code_to_prepare_porpoise_response_data-2020-10-12.R 7_Thompson_BOWL_cMMMP_ADD_experiment_data_2020-10-12 8_Thompson_BOWL_cMMMP_Distances_between_ADD_tests_&_CPODs_data_2020-10-12.txt 9_Thompson_BOWL_cMMMP_CPOD_porpoise_ClickCounts_data_2020-10-12 10_Thompson_BOWL_cMMMP_R_code_to_analyse_porpoise_responses_to_ADD_2020-10-12.R 11_Thompson_BOWL_cMMMP_Porpoise_responses_to_ADD_experiments_data_2020-10-12.txt 12_Thompson_BOWL_cMMMP_R_code_to_prepare_porpoise_return_times_data_2020-10-12.R 13_Thompson_BOWL_cMMMP_CPOD_porpoise_ClickDetails_data_2020-10-12.zip 14_Thompson_BOWL_cMMMP_expt_return_times_CPOD_data_2020-10-12.txt 15_Thompson_BOWL_cMMMP_baseline_return_times_CPOD_data_2020-10-12.txt 16_Thompson_BOWL_cMMMP_R_code_to_analyse_porpoise_return_times_after_ADD_playbacks_2020-10-12.R Piling noise and Acoustic Deterrent device signals were measured using calibrated broadband noise recorders (Soundtrap ST300HF, Ocean Instruments) moored 2 m above the seabed. An array of moored echolocation detectors (V.0 and V.1 CPODs (www.chelonia.co.uk)) was used to assess variation in harbour porpoise detections in relation to experimental ADD exposure. Data on the timing of piling operations, pile-hammer energies used, and pile penetration depths were provided by the wind farm developer (Beatrice Offshore Wind Ltd).

Mitigation measures to disperse marine mammals prior to pile-driving include acoustic deterrent devices and piling soft starts, but their efficacy remains uncertain. We developed a self-contained portable hydrophone cluster to detect small cetacean movements. Using an array of clusters within 10 km of foundation pile installations, we tested the hypothesis that harbour porpoises (Phocoena phocoena) respond to mitigation measures at offshore windfarm sites by moving away. During baseline periods, porpoise movements were evenly distributed in all directions. In contrast, animals showed significant directional movement away from sound sources during acoustic deterrent device use and piling soft starts. We demonstrate that porpoises respond to measures aimed to mitigate the most severe impacts of construction at offshore windfarms by swimming directly away from these sound sources. Portable directional hydrophone clusters now provide opportunities to characterise responses to disturbance sources across a broad suite of habitats and contexts. A full description of the data identifiers, R code and the data files required to repeat each analysis is provided in the text file: README_Graham_MEOW_cMMMP_Data_Packages.txt Data and code consist of the following 13 files/file packages. Graham_MEOW_cMMMP_construction_noise_data_2022-07-19.txt Graham_MEOW_cMMMP_construction_activity_data_2022-01-06.txt Graham_MEOW_cMMMP_directional_hydrophone_cluster_locations_and_distances_to_construction_data_2022-01-06.txt Graham_MEOW_cMMMP_R_code_to characterize_construction_noise_levels_2022-07-19.R Graham_MEOW_cMMMP_harbour_porpoise_click_data_2022-01-06.txt Graham_MEOW_cMMMP_R_code_to_process_and_analyse_harbour_porpoise_click_data_2022-07-19.R Depl1062.zip Depl1064.zip Depl1065.zip Depl1066.zip Depl1067.zip Depl1068.zip Depl1069.zip Funding provided by: Moray Offshore Wind Farm (East) Ltd.*Crossref Funder Registry ID: Award Number: Harbour porpoise clicks and noise levels during mitigation activities and baseline periods were recorded using tetrahedral clusters of hydrophones connected to four-channel underwater acoustic recorders (SoundTrap ST4300HF, Ocean Instruments NZ) deployed on the seabed. Recordings on each hydrophone cluster were processed in open-source PAMGuard software to determine received noise levels and classify harbour porpoise clicks. Data on the timing of construction activities were provided by the windfarm developer (Moray Offshore Wind Farm (East) Ltd).

Increasing levels of anthropogenic underwater noise have caused concern over their potential impacts on marine life. Offshore renewable energy developments and seismic exploration can produce impulsive noise which is especially hazardous for marine mammals because it can induce auditory damage at shorter distances and behavioral disturbance at longer distances. However, far-field effects of impulsive noise remain poorly understood, causing a high level of uncertainty when predicting the impacts of offshore energy developments on marine mammal populations. Here we used a 10-year dataset on the occurrence of coastal bottlenose dolphins over the period 2009–2019 to investigate far-field effects of impulsive noise from offshore activities undertaken in three different years. Activities included a 2D seismic survey and the pile installation at two offshore wind farms, 20–75 km from coastal waters known to be frequented by dolphins. We collected passive acoustic data in key coastal areas and used a Before-After Control-Impact design to investigate variation in dolphin detections in areas exposed to different levels of impulsive noise from these offshore activities. We compared dolphin detections at two temporal scales, comparing years and days with and without impulsive noise. Passive acoustic data confirmed that dolphins continued to use the impact area throughout each offshore activity period, but also provided evidence of short-term behavioral responses in this area. Unexpectedly, and only at the smallest temporal scale, a consistent increase in dolphin detections was observed at the impact sites during activities generating impulsive noise. We suggest that this increase in dolphin detections could be explained by changes in vocalization behavior. Marine mammal protection policies focus on the near-field effects of impulsive noise; however, our results emphasize the importance of investigating the far-field effects of anthropogenic disturbances to better understand the impacts of human activities on marine mammal populations.

Increasing levels of anthropogenic underwater noise have caused concern over their potential impacts on marine life. Offshore renewable energy developments and seismic exploration can produce impulsive noise which is especially hazardous for marine mammals because it can induce auditory damage at shorter distances and behavioural disturbance at longer distances. However, far-field effects of impulsive noise remain poorly understood, causing a high level of uncertainty when predicting the impacts of offshore energy developments on marine mammal populations. Here we used a 10-year dataset on the occurrence of coastal bottlenose dolphins over the period 2009-2019 to investigate far-field effects of impulsive noise from offshore activities undertaken in three different years. Activities included a 2D seismic survey and the pile installation at two offshore wind farms, 20-75 km from coastal waters known to be frequented by dolphins. We collected passive acoustic data in key coastal areas and used a Before-After Control-Impact design to investigate variation in dolphin detections in areas exposed to different levels of impulsive noise from these offshore activities. We compared dolphin detections at two temporal scales, comparing years and days with and without impulsive noise. Passive acoustic data confirmed that dolphins continued to use the impact area throughout each offshore activity period, but also provided evidence of short-term behavioural responses in this area. Unexpectedly, and only at the smallest temporal scale, a consistent increase in dolphin detections was observed at the impact sites during activities generating impulsive noise. We suggest that this increase in dolphin detections could be explained by changes in vocalization behaviour. Marine mammal protection policies focus on the near-field effects of impulsive noise; however, our results emphasize the importance of investigating the far-field effects of anthropogenic disturbances to better understand the impacts of human activities on marine mammal populations. Echolocation detectors (CPODs; Chelonia Ltd) were deployed between 2009 and 2019 to investigate the variation in dolphin detections in relation to the impulsive noise from three energy developments: a seismic survey for oil and gas exploration and the installation of foundation piles for two offshore wind farms (Beatrice Offshore Wind Farm and Moray East Offshore Wind Farm). Data on the timing of the seismic survey and piling operations were provided by the developers (Oil and Gas UK Ltd., COWRIE, Beatrice Offshore Wind Ltd. and Moray Offshore Wind Farm East). Data consist of 7 files and include the datasets and R code required to repeat all the analyses. A full description of the files provided in the Readme.txt file: OFB_FarField_DPH.csv OFB_FarField_BOWL.csv OFB_FarField_MEOW.csv OFB_FarField_BACI_Obtain_DPH_Dataset.R OFB_FarField_DPH_for_BACI.csv OFB_FarField_BACI_DPH_Models.R OFB_FarField_Readme.txt

Mitigation measures to disperse marine mammals prior to pile-driving include acoustic deterrent devices and piling soft starts, but their efficacy remains uncertain. We developed a self-contained portable hydrophone cluster to detect small cetacean movements. Using an array of clusters within 10 km of foundation pile installations, we tested the hypothesis that harbour porpoises (Phocoena phocoena) respond to mitigation measures at offshore windfarm sites by moving away. During baseline periods, porpoise movements were evenly distributed in all directions. In contrast, animals showed significant directional movement away from sound sources during acoustic deterrent device use and piling soft starts. We demonstrate that porpoises respond to measures aimed to mitigate the most severe impacts of construction at offshore windfarms by swimming directly away from these sound sources. Portable directional hydrophone clusters now provide opportunities to characterise responses to disturbance sources across a broad suite of habitats and contexts. A full description of the data identifiers, R code and the data files required to repeat each analysis is provided in the text file: README_Graham_MEOW_cMMMP_Data_Packages.txt Data and code consist of the following 13 files/file packages. Graham_MEOW_cMMMP_construction_noise_data_2022-07-19.txt Graham_MEOW_cMMMP_construction_activity_data_2022-01-06.txt Graham_MEOW_cMMMP_directional_hydrophone_cluster_locations_and_distances_to_construction_data_2022-01-06.txt Graham_MEOW_cMMMP_R_code_to characterize_construction_noise_levels_2022-07-19.R Graham_MEOW_cMMMP_harbour_porpoise_click_data_2022-01-06.txt Graham_MEOW_cMMMP_R_code_to_process_and_analyse_harbour_porpoise_click_data_2022-07-19.R Depl1062.zip Depl1064.zip Depl1065.zip Depl1066.zip Depl1067.zip Depl1068.zip Depl1069.zip Harbour porpoise clicks and noise levels during mitigation activities and baseline periods were recorded using tetrahedral clusters of hydrophones connected to four-channel underwater acoustic recorders (SoundTrap ST4300HF, Ocean Instruments NZ) deployed on the seabed. Recordings on each hydrophone cluster were processed in open-source PAMGuard software to determine received noise levels and classify harbour porpoise clicks. Data on the timing of construction activities were provided by the windfarm developer (Moray Offshore Wind Farm (East) Ltd).

Abstract1. Offshore windfarms require construction procedures that minimize impacts on protected marine mammals. Uncertainty over the efficacy of existing guidelines for mitigating near‐field injury when pile‐driving recently resulted in the development of alternative measures, which integrated the routine deployment of acoustic deterrent devices (ADD) into engineering installation procedures without prior monitoring by marine mammal observers.2. We conducted research around the installation of jacket foundations at the UK's first deep‐water offshore windfarm to address data gaps identified by regulators when consenting this new approach. Specifically, we aimed to (a) measure the relationship between noise levels and hammer energy to inform assessments of near‐field injury zones and (b) assess the efficacy of ADDs to disperse harbour porpoises from these zones.3. Distance from piling vessel had the biggest influence on received noise levels but, unexpectedly, received levels at any given distance were highest at low hammer energies. Modelling highlighted that this was because noise from pin pile installations was dominated by the strong negative relationship with pile penetration depth with only a weak positive relationship with hammer energy.4. Acoustic detections of porpoises along a gradient of ADD exposure decreased in the 3‐h following a 15‐min ADD playback, with a 50% probability of response within 21.7 km. The minimum time to the first porpoise detection after playbacks was > 2 h for sites within 1 km of the playback.5. Our data suggest that the current regulatory focus on maximum hammer energies needs review, and future assessments of noise exposure should also consider foundation type. Despite higher piling noise levels than predicted, responses to ADD playback suggest mitigation was sufficiently conservative. Conversely, strong responses of porpoises to ADDs resulted in far‐field disturbance beyond that required to mitigate injury. We recommend that risks to marine mammals can be further minimized by (1) optimizing ADD source signals and/or deployment schedules to minimize broad‐scale disturbance; (2) minimizing initial hammer energies when received noise levels were highest; (3) extending the initial phase of soft start with minimum hammer energies and low blow rates.

Readme.txt file for data packages associated with paper on “Harbour porpoise responses to pile-driving diminish over time”Description of data packagesGraham_BOWL_cMMMP_Data_Packages_Readme_2019-05-01.txtR code to prepare porpoise responses & construction activity dataR code to prepare the data file for analyses of porpoise responses to pile-driving, acoustic deterrent device (ADD) use and vessel activity. Filename = "Graham_BOWL_cMMMP_R_code_to_prepare_porpoise_response_data_2019-01-18.R"Graham_BOWL_cMMMP_R_code_to_prepare_porpoise_response_data_2019-05-01.RZip folder with 100 data files on porpoise occurrenceZip folder with 100 CPOD data files on porpoise occurrence for analyses of responses of porpoises to construction activity. Filename = "Graham_BOWL_cMMMP_CPOD_porpoise_ClickCounts_data_2019-01-18.zip” Each file has the data from a single CPOD deployment, cropped to remove data up to 23:59 on the day of deployment and from 00:00 on the day of retrieval. Each file has five columns = File, ChunkEnd, Nfiltered, Nall, MinsOn File is the filename of the original CPOD data file. ChunkEnd is the time in minutes. Nfiltered is the number of high and moderate quality porpoise clicks for each minute. Nall is the total number of unfiltered clicks for each minute. MinsOn indicates whether the CPOD was recording during that minute (1) or not (0).Graham_BOWL_cMMMP_CPOD_porpoise_ClickCounts_data_2019-01-18.zipData on piling activity by turbine/OTM locationData on the piling activity summarised by turbine/OTM location (see Table S1). Filename = “Graham_BOWL_cMMMP_Piling_summary_by_turbine_data_2019-01-18.csv” Twelve columns = turbine, start_time, end_time, piling_duration, piling_duration_hours, total_blow_count, max_energy, interval_days, daystocomplete, piling_order, lat, long. Turbine is the code for the turbine/OTM location. Start_time is the time when piling of the first pile started. End_time is the time when piling of the fourth pile ended. Piling_duration is the duration of active piling in minutes. Piling_duration_hours is the duration of active piling in hours. Total_blow_count is the total number of hammer blows. Max_energy is the maximum hammer energy (kJ). Interval_days is the time difference (in days) between the end of piling at the previous location and the start of piling at the current location. Daystocomplete is the number of days from the start to the end of piling. Piling_order is the order in which locations were piled, equivalent to the cumulative number of locations piled at the end of piling at that location. Lat is the latitude in decimal degrees. Long is the longitude in decimal degrees.Graham_BOWL_cMMMP_Piling_summary_by_turbine_data_2019-01-18.csvData on piling activity by turbine/OTM location & dayData on the piling activity summarised by turbine/OTM location and day. Filename = “Graham_BOWL_cMMMP_Piling_summary_by_day_data_2019-01-18.csv” Twenty-two columns = julian, start_time, end_time, total_duration, total_duration_hours, total_blow_count, max_energy, no_turbines_piled, turbine1, start_time1, end_time1, duration1, duration_hours1, blow_count1, max_energy1, turbine2, start_time2, end_time2, duration2, duration_hours2, blow_count2, max_energy2.Graham_BOWL_cMMMP_Piling_summary_by_day_data_2019-01-18.csvData on CPOD deploymentsData on the CPOD deployments. Filename = “Graham_BOWL_cMMMP_POD_deployment_data_2019-01-18.csv” Eight columns = Dep_no, POD_number, Deployment_date, Location_ID, Latitude, Longitude, Retrieval_date, Data_end_date. Dep_no is the unique number assigned to each CPOD deployment. POD_number is the unique individual number for the CPOD used in that deployment. Deployment_date is the date when that CPOD was deployed at sea. Location_ID is the numeric identifier for the location at which that CPOD was deployed. Latitude is the latitude in decimal degrees for that deployment. Longitude is the longitude in decimal degrees for that deployment Retrieval_date is the date when that CPOD was retrieved at sea. Data_end_date is the date when the data on the CPOD’s SD card ended.Graham_BOWL_cMMMP_POD_deployment_data_2019-01-18.csvData on AIS detections within 1km of CPODsData on the Automatic Identification System (AIS) vessel detections within 1km of each CPOD. Raw AIS data for 2017 for the Moray Firth were supplied to BOWL by Astra Paging Ltd., Sliven, Bulgaria. This AIS data consisted of the location (latitude and longitude) of each AIS vessel at 5 minutes intervals. These data were processed in R to select only those AIS vessel locations within 1 km of each CPOD when the CPOD was operational (i.e. deployed and recording data). Filename = “Graham_BOWL_cMMMP_5min_boat_withineach_1kmCPODbuffer _data_2019-01-18.csv” Five columns = MMMSI.x, DATE.TIME.x, Deployment.number, BOAT.LATITUDE, BOAT.LONGITUDE. MMMSI.x is the unique 9 digit vessel identifier. DATE.TIME.x is the time of the vessel detection. Deployment.number is the CPOD deployment to which that vessel detection refers. BOAT.LATITUDE is the latitude in decimal degrees for that vessel detection. BOAT.LONGITUDE is the longitude in decimal degrees for that vessel detection.Graham_BOWL_cMMMP_5min_boat_withineach_1kmCPODbuffer_data_2019-01-18.csvData on AIS detections within 500m of CPODsData on the number of AIS vessel detections within 500m of each CPOD. Raw AIS data for 2017 for the Moray Firth were supplied to BOWL by Astra Paging Ltd., Sliven, Bulgaria. This AIS data consisted of the location (latitude and longitude) of each AIS vessel at 5 minute intervals. These data were processed in R to select only AIS vessel locations within 500 m of each CPOD when the CPOD was operational (i.e. deployed and recording data). Filename = “Graham_BOWL_cMMMP_5min_boat_withineach_500mCPODbuffer _data_2019-01-18.csv” Five columns = MMMSI.x, DATE.TIME.x, Deployment.number, BOAT.LATITUDE, BOAT.LONGITUDE. MMMSI.x is the unique 9 digit vessel identifier. DATE.TIME.x is the time of the vessel detection. Deployment.number is the CPOD deployment to which that vessel detection refers. BOAT.LATITUDE is the latitude in decimal degrees for that vessel detection. BOAT.LONGITUDE is the longitude in decimal degrees for that vessel detection.Graham_BOWL_cMMMP_5min_boat_withineach_500mCPODbuffer_data_2019-01-18.csvData on the predicted received sound exposure levelsData on the predicted received sound exposure levels (SEL; dB re 1 µPa^2 s), single pulse for an impact strike with the maximum hammer energy for that piling location, depth averaged for each CPOD sampling sites. Filename = “Graham_BOWL_cMMMP_Received_noise_levels_18pilinglocations_allCPODlocations_data_2019-05-01.csv” Six columns = Deployment_number, Turbine, Unweighted_SS_SEL, NOAA_SS_SEL, Southall_SS_SEL, Aud_SS_SEL. Deployment_number is the CPOD deployment number for that location. Turbine is the piling location. Unweighted_SS_SEL is the unweighted the predicted received sound exposure level (SEL; dB re 1 µPa^2 s), single pulse for an impact strike with the maximum hammer energy for that piling location, depth averaged for that CPOD site. NOAA_SS_SEL is the predicted SEL frequency weighted by the generalised weighting function for high frequency cetaceans proposed by NOAA (2016 & 2018: see methods in paper for references). Southall_SS_SEL is the predicted SEL frequency weighted by the high-frequency cetacean weighting function proposed by Southall et al. (2007: see methods in paper for reference). Aud_SS_SEL is the predicted SEL frequency weighted by a harbour porpoise specific audiogram from Kastelein et al. (2010: see methods in paper for reference).Graham_BOWL_cMMMP_Received_noise_levels_18pilinglocations_allCPODlocations_data_2019-05-01.csvR code to analyse porpoise responses to construction activityR code to analyse porpoise responses to pile-driving, acoustic deterrent device (ADD) use and vessel activity. Filename = "Graham_BOWL_cMMMP_R_code_to_analyse_porpoise_responses_2019-01-18.R"Graham_BOWL_cMMMP_R_code_to_analyse_porpoise_responses_2019-05-01.RData on porpoise responses (from CPODs) for analyses of responses of porpoises to piling, acoustic deterrent device use and vessel activitFilename = “Graham_BOWL_cMMMP_Porpoise_responses_to_construction_data_2019-05-01.csv” Twenty-four columns = dep_no, turbine, dph24, dph12, base24, base12, prop24, prop12, resp24_50, resp12_50, location, distance, vessels24_1km, vessels12_1km, vessels24_500m, vessels12_500m, duration, piling_order, ADD, pod, Unweighted_SS_SEL, NOAA_SS_SEL, Southall_SS_SEL, Aud_SS_SEL. dep_no is the CPOD deployment number. turbine is the piling location. dph24 is the number of detection positive hours in the 24-hour period from the end of piling. dph12 is the number of detection positive hours in the 12-hour period from the end of piling. base24 is the number of detection positive hours in the 24-hour baseline period. base12 is the number of detection positive hours in the 12-hour baseline period. prop24 is (dph24 – base24)/base24. prop12 (dph12 – base12)/base12. resp24_50 is the 24-hour porpoise response coded as 0 or 1: if prop24 <= -0.5, response = 1. resp12_50 is the 12-hour porpoise response coded as 0 or 1: if prop12 <= -0.5, response = 1. location is the numeric identifier for the location at which that CPOD was deployed. distance is the distance to piling (km). vessels24_1km is the number of AIS vessel detections in the 24-h response period within 1km of the CPOD. vessels12_1km is the number of AIS vessel detections in the 12-h response period within 1km of the CPOD. vessels24_500m is the number of AIS vessel detections in the 24-h response period within 500m of the CPOD. vessels12_500m is the number of AIS vessel detections in the 12-h response period within 500m of the CPOD. duration is the duration of active piling in hours. piling_order is the order in which locations were piled, equivalent to the cumulative number of locations piled at the end of piling at that location. ADD is whether or not ADD mitigation was used prior to piling, coded “Y” (Yes) or “N” (No). pod is the CPOD number. Unweighted_SS_SEL, NOAA_SS_SEL, Southall_SS_SEL, Aud_SS_SEL are the predicted received sound exposure levels as above for data file 8.Graham_BOWL_cMMMP_Porpoise_responses_to_construction_data_2019-05-01.csvData on predicted porpoise responses for the model of 24-h response with distanceData on predicted porpoise responses for the model (24-h response ~ log(distance)*piling order + no. vessel locations_1km; see Table 1, Model a) used in the R code in the ESM to calculate the number of individuals displaced using behavioural-response curves. Filename = “Graham_BOWL_cMMMP_bootstrapped_predictions_m8_24_data_2019-05-01.csv” Six columns = distance, zorder, zvessels_1km, fit, lci, uci distance is the distance to piling (km). zorder is the standardised value of piling order (order in which the locations were piled): values were standardised by subtracting the mean piling order and dividing by the standard deviation of the mean. zvessels_1km is the standardised value of the number of AIS vessel detections in the 24-h response period within 1km of the CPOD: values were standardised by subtracting the mean number of AIS vessel detections and dividing by the standard deviation of the mean. fit are the predicted values of the probability of response estimated using bootstrapping. lci are the predicted values of the lower 95 % confidence interval of the probability of response estimated using bootstrapping. uci are the predicted values of the upper 95 % confidence interval of the probability of response estimated using bootstrapping.Graham_BOWL_cMMMP_bootstrapped_predictions_m8_24_data_2019-05-01.csv Estimating impacts of offshore windfarm construction on marine mammals requires data on displacement in relation to different noise levels and sources. Using echolocation detectors and noise recorders, we investigated harbour porpoise behavioural responses to piling noise during the 10-month foundation installation of a North Sea windfarm. Current UK guidance assumes total displacement within 26 km of pile driving. In contrast, we recorded a 50 % probability of response within 7.4 km (95 % CI = 5.7 – 9.4) at the first location piled, decreasing to 1.3 km (95 % CI = 0.2 – 2.8) by the final location; representing 28 % (95 % CI = 21 – 35) and 18 % (95 % CI = 13 – 23) displacement of individuals within 26 km. Distance proved as good a predictor of responses as audiogram weighted received levels, presenting a more practicable variable for environmental assessments. Critically, acoustic deterrent device (ADD) use and vessel activity increased response levels. Policy and management to minimise impacts of renewables on cetaceans have concentrated on pile-driving noise. Our results highlight the need to consider trade-offs between efforts to reduce far-field behavioural disturbance and near-field injury through ADD use.

Increasing levels of anthropogenic underwater noise have caused concern over their potential impacts on marine life. Offshore renewable energy developments and seismic exploration can produce impulsive noise which is especially hazardous for marine mammals because it can induce auditory damage at shorter distances and behavioural disturbance at longer distances. However, far-field effects of impulsive noise remain poorly understood, causing a high level of uncertainty when predicting the impacts of offshore energy developments on marine mammal populations. Here we used a 10-year dataset on the occurrence of coastal bottlenose dolphins over the period 2009-2019 to investigate far-field effects of impulsive noise from offshore activities undertaken in three different years. Activities included a 2D seismic survey and the pile installation at two offshore wind farms, 20-75 km from coastal waters known to be frequented by dolphins. We collected passive acoustic data in key coastal areas and used a Before-After Control-Impact design to investigate variation in dolphin detections in areas exposed to different levels of impulsive noise from these offshore activities. We compared dolphin detections at two temporal scales, comparing years and days with and without impulsive noise. Passive acoustic data confirmed that dolphins continued to use the impact area throughout each offshore activity period, but also provided evidence of short-term behavioural responses in this area. Unexpectedly, and only at the smallest temporal scale, a consistent increase in dolphin detections was observed at the impact sites during activities generating impulsive noise. We suggest that this increase in dolphin detections could be explained by changes in vocalization behaviour. Marine mammal protection policies focus on the near-field effects of impulsive noise; however, our results emphasize the importance of investigating the far-field effects of anthropogenic disturbances to better understand the impacts of human activities on marine mammal populations. Echolocation detectors (CPODs; Chelonia Ltd) were deployed between 2009 and 2019 to investigate the variation in dolphin detections in relation to the impulsive noise from three energy developments: a seismic survey for oil and gas exploration and the installation of foundation piles for two offshore wind farms (Beatrice Offshore Wind Farm and Moray East Offshore Wind Farm). Data on the timing of the seismic survey and piling operations were provided by the developers (Oil and Gas UK Ltd., COWRIE, Beatrice Offshore Wind Ltd. and Moray Offshore Wind Farm East). Data consist of 7 files and include the datasets and R code required to repeat all the analyses. A full description of the files provided in the Readme.txt file: OFB_FarField_DPH.csv OFB_FarField_BOWL.csv OFB_FarField_MEOW.csv OFB_FarField_BACI_Obtain_DPH_Dataset.R OFB_FarField_DPH_for_BACI.csv OFB_FarField_BACI_DPH_Models.R OFB_FarField_Readme.txt