Michael R. Schirmacher Manuela M. Huso Michael Whitby Cris D. Hein Jeffrey Clerc 20250113 tabular digital data Online U.S. Geological Survey https://doi.org/10.5066/P14NMJJR Jeffrey Clerc Manuela M. Huso Michael R. Schirmacher Michael Whitby Cris D. Hein 2025 publication Online PLoS ONE Wind energy is important for achieving net-zero greenhouse gas emissions but also contributes to global bat mortality. Current strategies to minimize bat mortality due to collision with wind-turbine blades fall broadly into two categories: curtailment (limiting turbine operation during high-risk periods) and deterrence (discouraging bat activity near turbines). Recently, there has been interest in combining these strategies to achieve greater reductions in bat fatalities than either strategy might achieve in isolation. To investigate the effectiveness of combining curtailment with ultrasonic deterrent minimization strategies, we deployed six ultrasonic deterrents at nacelle height on 16 experimental turbines at Avangrid Renewables' Blue Creek Wind Energy Facility. We rotated between four conditions (normal operations, curtailment only, deterrent only, curtailment and deterrent) randomly assigned to four wind turbines each night between 15 June and 3 October 2017. We searched daily for bat carcasses and associated those that were believed to have been killed the previous night with the treatment that was applied to that turbine on the previous night. The data set reflects mortality counts at each turbine on each night, and what the meteorological conditions were on each night. We also summarized turbine blade rotation over the entire night. To investigate the effectiveness of combining curtailment with ultrasonic deterrent minimization strategies in decreasing bat fatalities at wind energy facilities. It is critical that the experimental design be incorporated in any analysis. The turbines acted as the statistical blocking factor. All treatments were assigned to an equal number of nights (28 each) over the study period and to an equal number of turbines (4 each) on each night. A simple summary of mortality in each treatment without accounting for the blocking factors will be incorrect. 20170614 20171003 Ground condition (field visits); turbine-specific data for wind speed and blade rotation Complete None planned Paulding and Van Wert Counties, Ohio, USA -84.6799 -84.3723 41.0794 40.8117 ISO 19115 Topic Category biota environment USGS Thesaurus wind energy bats wildlife habitat alteration and disturbance land use change mathematical modeling None bat mortality wildlife mortality USGS Metadata Identifier USGS:675caee6d34e189836d1d966 Common geographic areas Ohio Integrated Taxonomic Information System (ITIS), accessed 20241213 Lasiurus borealis Lasiurus cinereus Lasionycteris noctivagans Eptesicus fuscus Nycticeius humeralis Integrated Taxonomic Information System (ITIS) 20241212 Comma Separated Value File Washington, D.C. Integrated Taxonomic Information System (ITIS) http://itis.gov We use Lasiurus cinereus to denote the hoary bat, but Aeroestes cinereus refers to the same species in the taxonomic hierarchy. Expertise Kingdom Animalia animals Subkingdom Bilateria triploblasts Infrakingdom Deuterostomia Phylum Chordata chordates Subphylum Vertebrata vertebrates Infraphylum Gnathostomata Superclass Tetrapoda Class Mammalia mammals Subclass Theria Infraclass Eutheria Order Chiroptera bats Suborder Yangochiroptera Superfamily Vespertilionoidea Family Vespertilionidae vesper bats vespertilionid bats evening bats Subfamily Vespertilioninae Tribe Lasiurini hairy-tailed bats Genus Lasiurus hairy-tailed bats red bats Species Lasiurus borealis red bat Eastern Red Bat TSN: 180016 Genus Aeorestes hoary bats Species Aeorestes cinereus North American Hoary Bat Tribe Eptesicini serotines Genus Lasionycteris silver-haired bats Species Lasionycteris noctivagans Silver-haired Bat Black Bat Silver-black Bat TSN: 180014 Genus Eptesicus big brown bats Subgenus Eptesicus (Eptesicus) Species Eptesicus fuscus Big Brown Bat TSN: 180008 Tribe Nycticeiini evening bats Genus Nycticeius evening bats Species Nycticeius humeralis Evening Bat North American Evening Bat TSN: 180022 No access constraints. No use constraints. Public domain data from the U.S. Government are freely redistributable with proper metadata and source attribution. The U.S. Geological Survey requests to be acknowledged as originator of these data in future products or derivative research. Users are advised to read the dataset's metadata thoroughly to understand appropriate use and data limitations. Forest and Rangeland Ecosystem Science Center Data Manager mailing and physical

777 NW 9th Street, Suite 400

Corvallis Oregon 97330 USA 541-750-1030 fresc_outreach@usgs.gov Avangrid Renewables The R Foundation for Statistical Computing 2024 publication Because the turbines represented the statistical blocking factor and detection probability was assumed constant within turbine, we used simple count as an index of mortality. We fit the data to statistical models using the R package "glmmTMB". www.r-project.org Throughout the study NRG Systems monitored the functional status of the deterrents on a nightly basis using a cellular modem connection and noted whether there were known deterrent malfunctions at a wind turbine on any given night. We also examined 10-minute average RPM of each wind turbine in relation to 10-minute average wind speed throughout each night to determine if the curtailment was implemented as assigned. Any deviations were recorded. We analyzed results based on the "effective treatment," indicating how the turbine and/or deterrent actually operated on a given night regardless of its assigned treatment. Searchers walked along 5-m-wide transects within a 90-m radius of each experimental wind turbine. We selected relatively large plot sizes to address two potential sources of detection bias: 1) undercounting fatalities that fall farther from the wind turbine when struck near the blade tip, which could lead to an overestimate of the effectiveness of ultrasonic deterrents, and 2) undercounting fatalities that fall farther from the wind turbine when struck on windier nights, which could lead to an overestimate of the effectiveness of the curtailment treatment. All data are confirmed to be within ranges, without duplication or omissions. We excluded bat carcasses that we were not able to unequivocally associate with treatment and meteorological conditions on the night they were killed. During the 112-night study period, we found 234 bats at experimental wind turbines, 154 of which we considered to be "fresh," i.e., thought to have been killed the previous night and therefore associated with the treatment conditions of the previous night. We excluded 80 carcasses from our analysis. Our experiment ran for 1,792 turbine-nights (16 turbines * 112 nights) but our final dataset comprised 1,727 turbine-nights. We could not conduct 62 (3.5%) planned searches due to weather, pesticide applications, or other uncontrollable factors and we excluded 3 turbine-nights because wind turbines were down for repairs and not functioning, or RPM data were not recorded. On 20 of the 1,727 turbine-nights (1%) the wrong treatment was implemented. Such misalignments can occur when wind turbines assigned to curtail do not receive the signal to feather the blades below the curtailment wind speed or wind turbines assigned to not curtail may erroneously receive a signal to curtail. Not applicable. Not applicable. Study design: We selected 16 wind turbines for the study at Blue Creek with two constraints: 1) landowner approval was received and 2) each turbine was >250 m from the nearest wind turbine. Landowners were compensated for crop loss in the 90-m-radius circle (2.5 ha) cleared beneath each wind turbine in which carcass searches were conducted daily. A 250-m turbine separation distance ensured that all bat fatalities discovered could be clearly associated with the experimental wind turbine and not a neighboring one. From earlier studies, we expected relatively high variation in mortality among turbines. We conducted a pre-study statistical power analysis using data collected from earlier studies. The analysis indicated that for our 2x2 factorial experiment, a randomized block design to control variation in mortality among wind turbines would better detect treatment differences than a completely randomized design. On each night, each treatment was randomly assigned to four wind turbines. Assignment was re-randomized nightly with the constraint that the treatment assigned to each turbine was balanced (four nights of each treatment) every 16 nights. At the beginning of the study the wind energy facility operators were given a schedule of which wind turbines to curtail on which nights. Treatments were implemented nightly from 1 hr prior to sunset to 1 hr post sunrise for 112 nights, from 14 June through 2 Oct 2017. Daily searches for carcasses at each experimental wind turbine began 15 June and ended 3 October 2017. We selected this time period based on fatality data collected at the site in 2012 and 2013 that indicated this time interval represented the period during which the most fatalities occurred. All turbines were searched by human searchers every day. Searchers walked along 5-m-wide transects within a 90-m radius of each experimental wind turbine. For each study night we summarized the average wind speed and the proportion of the night (from data measured in 10-minute intervals) during which RPM was more than 2 for each wind turbine. We chose a cutoff value of 2 RPM to indicate when a wind turbine was curtailed because it represented a natural break in the data (fully operational wind turbines never fell below 3 RPM). We used 10-minute RPM vs. 10-minute wind speed graphs produced for each wind turbine to ascertain whether curtailment was appropriately applied, recoding treatment to effective treatment when disagreements occurred. We did the same for ultrasonic deterrents. We recorded a count of carcasses in each species group at each wind turbine on each night. Because the turbines represented the statistical blocking factor and detection probability was assumed constant within turbine, we used simple count as an index of mortality. We fit the data to statistical models using the R package "glmmTMB". 20171003 PLOSOne.modelData.csv Comma Separated Value (CSV) file containing data. Producer Defined day.of.year.night Julian Day (01.01.14 = 1) on which night begins. Producer Defined 164 275 Days searchDay Date on which search occurs. Producer Defined List of calendar dates in M/D/YYYY format. int Number of 10-minute intervals in night. Producer Defined 53 74 Interval turbine Turbine identifier. Producer Defined Alphanumeric text. trt Assigned treatment code. Producer Defined CD Curtailment and deterrent. Producer defined CO Curtailment only. Producer defined OD Deterrent only. Producer defined OO Control. Producer defined trtCurt Indicates whether curtailment was assigned. Producer Defined True Curtailment was assigned. Producer defined False Curtailment was not assigned. Producer defined trtDet Indicates whether deterrence was assigned. Producer Defined True Deterrence was assigned. Producer defined False Deterrence was not assigned. Producer defined effective Effective (implemented) treatment code. Producer Defined CD Curtailment and deterrent. Producer defined CO Curtailment only. Producer defined OD Deterrent only. Producer defined OO Control. Producer defined effCurt Indicates whether curtailment was implemented. Producer Defined True Curtailment was implemented. Producer defined False Curtailment was not implemented. Producer defined effDet Indicates whether deterrent was implemented. Producer Defined True Deterrent was implemented. Producer defined False Deterrent was not implemented. Producer defined all Total number of fresh bats found. "Fresh" describes bats thought to have been killed the previous night and therefore associated with the treatment conditions of the previous night. Producer Defined 0 3 individual bats labo Number of fresh eastern red bats (Lasiurus borealis) found. "Fresh" describes bats thought to have been killed the previous night and therefore associated with the treatment conditions of the previous night. Producer Defined 0 3 individual bats lowfreq Number of fresh low-frequency bats (hoary, silver-haired, and big brown bats) found. "Fresh" describes bats thought to have been killed the previous night and therefore associated with the treatment conditions of the previous night. Producer Defined 0 2 individual bats MET2.avgws Met Tower 2 average wind speed. Producer Defined 1.949847628 11.51934666 meters per second MET2.wgt5 Met Tower 2 proportion of 10-minute intervals in which average wind speed > 5 meters per second. Producer Defined 0.0 1.0 unitless Turb.propSpin Individual turbine proportion of 10-minute intervals during which turbine was spinning > 2 RPM. Producer Defined 0.0 1.0 unitless GS ScienceBase U.S. Geological Survey mailing address

Denver Federal Center, Building 810, Mail Stop 302

Denver CO 80225 United States 1-888-275-8747 sciencebase@usgs.gov Unless otherwise stated, all data, metadata and related materials are considered to satisfy the quality standards relative to the purpose for which the data were collected. Although these data and associated metadata have been reviewed for accuracy and completeness and approved for release by the U.S. Geological Survey (USGS), no warranty expressed or implied is made regarding the display or utility of the data for other purposes, nor on all computer systems, nor shall the act of distribution constitute any such warranty. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. CSV https://doi.org/10.5066/P14NMJJR None 20250113 Forest and Rangeland Ecosystem Science Center Data Manager mailing and physical

777 NW 9th Street, Suite 400

Corvallis Oregon 97330 USA 541-750-1030 fresc_outreach@usgs.gov FGDC Biological Data Profile of the Content Standard for Digital Geospatial Metadata FGDC-STD-001.1-1999