Auteur moral

Auteur secondaire

Résumé

"Wind turbines negatively affect bats through mortality, which can be exacerbated by attraction behaviours, and loss of habitat use caused by avoidance behaviours. However, potential mechanisms driving bat responses to wind turbines are still poorly understood. This is especially true of red aviation lighting, designed to prevent aircraft collisions and implemented in many countries, that could be perceived by bats from a long distance and lead to a response at a large spatial scale. We assessed the role of wind turbine red aviation lighting in the behavioural responses (attraction and avoidance) of bats. To this end, we acoustically quantified the activity of three functional bat guilds (long-, medium- and short-range echolocators) at three wind farms using a triplet sampling design: recordings were conducted simultaneously at wind turbines illuminated throughout the night, wind turbines equipped with the aircraft detection lighting system (ADLS) and illuminated an average of 12% of the night and (iii) control sites without nearbywind turbine. Thirteen and nine triplets were sampled at wooded edges ~250?m from the nearest wind turbine and in open habitats at the base of the turbine, respectively, during two consecutive nights in June 2021 in the Uckermark district (north-east Germany). We found that acoustic activity was higher overall at sites near wind turbines illuminated throughout the night than at control sites for all functional guilds and both at wooded edges and in open habitats, indicating local attraction behaviours towards wind turbines that might increase collision risks. Activity at sites near wind turbines with ADLS was lower overall than at sites near wind turbines illuminated throughout the night, and similar to control sites, suggesting that part-night lighting could contribute to reducing bat attraction towards wind turbines"

Editeur

Journal of applied ecology

Descripteur Urbamet

Descripteur écoplanete

faune sauvage ;éolienne ;impact sur l'environnement

Thème

Énergie - Climat ;Nature ;Méthodes - Techniques ;Sciences de la terre ;Risques

Texte intégral

J Appl Ecol. 2026;63:e70226.      ?|?1 of 13 https://doi.org/10.1111/1365-2664.70226 wileyonlinelibrary.com/journal/jpe Received: 23 April 2025? |?Accepted: 17 October 2025 DOI: 10.1111/1365-2664.70226   R E S E A R C H A R T I C L E Minimizing aviation lighting duration reduces bat attraction to wind turbines Gaëlle Larnoy1?| Fabien Verniest2 ?| Christian Kerbiriou2 ?| Isabelle Le Viol2 ?| Pauline Lefebvre1?| Nicolas Valet1?| Kévin Barré2,3 ?| Camille Leroux1,2 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2025 The Author(s). Journal of Applied Ecology published by John Wiley & Sons Ltd on behalf of British Ecological Society. Gaëlle Larnoy and Fabien Verniest contributed equally as first authors. Kévin Barré and Camille Leroux contributed equally as last authors. 1Auddicé Biodiversité?ZAC du Chevalement, Roost- Warendin, France 2Centre d'Ecologie et des Sciences de la Conservation (CESCO), Muséum National  d'Histoire Naturelle, Centre National  de la Recherche Scientifique, Sorbonne Université, Station de Biologie Marine,  Concarneau Cedex, France 3Complex Systems Group (NEXUS::CSR),  Faculty of Science, Technology, and Medicine (FSTM), University  of Luxembourg, Esch- sur- Alzette, Luxembourg Correspondence Fabien Verniest Email: fabien.verniest@mnhn.fr Funding information Auddicé biodiversité; Agence de la transition écologique; Association Nationale de la Recherche et de la  Technologie, Grant/Award Number:  2019/1566 Handling Editor: Silke Bauer Abstract 1. Wind turbines negatively affect bats through mortality, which can be exacerbated by attraction behaviours, and loss of habitat use caused by avoidance behaviours. However, potential mechanisms driving bat responses to wind turbines are still poorly understood. This is especially true of red aviation lighting, designed to prevent aircraft collisions and implemented in many countries, that could be perceived by bats from a long distance and lead to a response at a large spatial scale. 2. We assessed the role of wind turbine red aviation lighting in the behavioural re- sponses (attraction and avoidance) of bats. To this end, we acoustically quanti- fied the activity of three functional bat guilds (long- , medium- and short- range echolocators) at three wind farms using a triplet sampling design: recordings were conducted simultaneously at (i) wind turbines illuminated throughout the night, (ii) wind turbines equipped with the aircraft detection lighting system (ADLS) and illuminated an average of 12% of the night and (iii) control sites without nearby wind turbine. Thirteen and nine triplets were sampled at wooded edges ~250?m  from the nearest wind turbine and in open habitats at the base of the turbine, re- spectively, during two consecutive nights in June 2021 in the Uckermark district (north- east Germany). 3. We found that acoustic activity was higher overall at sites near wind turbines il- luminated throughout the night than at control sites for all functional guilds and both at wooded edges and in open habitats, indicating local attraction behaviours towards wind turbines that might increase collision risks. 4. Activity at sites near wind turbines with ADLS was lower overall than at sites near wind turbines illuminated throughout the night, and similar to control sites, suggesting that part- night lighting could contribute to reducing bat attraction to- wards wind turbines. 5. Synthesis and applications. This study provides empirical evidence that attrac- tion behaviour of bats towards wind turbines is driven, at least partially, by red https://doi.org/10.1111/1365-2664.70226 www.wileyonlinelibrary.com/journal/jpe mailto: https://orcid.org/0000-0001-5744-3185 https://orcid.org/0000-0001-6080-4762 https://orcid.org/0000-0003-3475-5615 https://orcid.org/0000-0001-5368-4053 https://orcid.org/0000-0002-4984-3485 http://creativecommons.org/licenses/by/4.0/ mailto:fabien.verniest@mnhn.fr http://crossmark.crossref.org/dialog/?doi=10.1111%2F1365-2664.70226&domain=pdf&date_stamp=2025-12-09 2 of 13? |? ?? LARNOY et al. 1? |? INTRODUC TION Mitigating climate change requires a major reduction in greenhouse  gas emissions (IPCC, 2022). One of the main levers to achieve this  reduction is the development of renewable energy sources, such as wind energy, hydropower and solar energy. Among these, wind en- ergy is playing an increasingly key role (IRENA, 2022), with electric- ity production from wind turbines projected to double in just a few years at the global scale (Hutchinson & Zhao, 2023). However, many articles have reported negative impacts of wind energy production on the environment (Saidur et al., 2011; Wang & Wang, 2015), in- cluding biodiversity. This leads to a conflict between climate change mitigation and biodiversity conservation: the so- called green- green dilemma (Voigt et al., 2019). Extensive research has shown that wind turbines can have nega- tive effects on airborne biodiversity, such as birds and bats (Kuvlesky et al., 2007; Schuster et al., 2015), which can contribute to the decline of species populations (Duriez et al., 2023; Frick et al., 2017). Consequences on bats have been particularly investigated and are twofold: (i) mortal- ity caused by collision (Arnett et al., 2008; Kunz et al., 2007; O'Shea  et al., 2016; Rydell et al., 2010), that might be exacerbated by attraction behaviours (Cryan et al., 2014; Ellerbrok et al., 2023; Horn et al., 2008; Richardson et al., 2021); and (ii) loss of habitat use caused by avoidance behaviours (Barré et al., 2018; Ellerbrok et al., 2022; Gaultier et al., 2023; Minderman et al., 2017). Whilst previous studies suggest that all bat foraging guilds can exhibit both attraction and avoidance behaviours towards wind turbines, the prevalence of one behaviour over the other might be attributed to many factors, such as the life cycle stage (Ellerbrok et al., 2022;  McKay  et  al.,  2024), the characteristics and operation of wind turbines (Cryan et al., 2014; Ellerbrok et al., 2024; Leroux et al., 2023, 2024), the distance from the tur- bines (Gaultier et al., 2023; Leroux et al., 2023) and the local hab- itat (Leroux et al., 2022; Reusch et al., 2022; Scholz et al., 2025; Sotillo et al., 2024). The identification of these multiple factors in recent investigations suggests the co- occurrence of several underlying mechanisms that are currently unknown. Therefore, very few opportunities with limited effectiveness in reducing at- traction and avoidance behaviours have been explored so far. For instance, the adequacy of the UNEP/EUROBATS recommendation  for the spatial positioning of wind turbines (distance to the nearest wooded edge >200?m; Rodrigues et al., 2015) has been questioned (Barré et al., 2018), in addition to being poorly implemented (Barré et al., 2022). Another widespread mitigation measure is wind tur- bine curtailment using blade feathering when bats are highly active and energy production is low. Although this measure successfully mitigates collision risks, its effectiveness remains highly variable (Adams et al., 2021; Whitby et al., 2024). Investigating the mecha- nisms underlying bat responses to wind turbines is therefore nec- essary to design new mitigation measures for the negative impacts of wind energy on bats. Various hypotheses, involving different sensory modalities and spatial scales, have been advanced in the literature (Cryan & Barclay, 2009; Guest et al., 2022; Jonasson et al., 2024). For instance, attraction at small spatial scales could be due to the high density of insects at wind turbines (Horn et al., 2008; McKay  et al., 2024; Voigt, 2021). Wind turbines may also attract bats be- cause they are perceived as potential trees for roosting or mating (Cryan, 2008). Conversely, airflow disturbance generated by mov- ing blades leads to avoidance behaviour downwind of wind tur- bines for some species, which could be attributed to poorer flight and foraging conditions (Leroux et al., 2024). Other mechanisms  underlying bat responses to wind turbines, such as the effects of red aviation lighting, which is designed to prevent collisions with aircraft, may play a significant role in bat responses to wind tur- bines (Voigt et al., 2018). However, this mechanism has received far too little attention despite the widely recognized effects of artificial  light  at  night  (ALAN)  on  bats  (Stone  et  al., 2015; Voigt et al., 2021). The red aviation lighting of wind turbines could be perceived by bats from a long distance, resulting in a response on a large spatial scale (Jonasson et al., 2024). Indeed, it may attract bats, as  suggested by Voigt et al. (2018) who described such behaviour in two migrating Pipistrellus species towards red lighting in coastal meadows  during  summer  migration.  In  contrast,  red  aviation  lighting may also result in avoidance behaviour, as suggested by Barré et al. (2021) who demonstrated that all bat guilds seek ref- uge in cluttered environments when exposed to red streetlighting during foraging. Managing the red aeronautical lighting of turbines  may consequently offer an easy- to- implement tool to mitigate the  negative  impacts  of  wind  turbines  on  bats.  Only  four  stud- ies,  all  conducted  in  North  America,  have  hitherto  investigated  aviation lighting. We also demonstrate that smart lighting of wind turbines, such as the ADLS, could cost- effectively help mitigate disruption of bat habitat use and the associated collision risks. Implementing adaptive lighting strategies could  therefore represent a practical step towards balancing wind energy development with bat conservation. K E Y W O R D S acoustic activity, ADLS, ALAN, avoidance, Chiroptera, local scale, onshore wind energy, smart  lighting 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense ??? |? 3 of 13LARNOY et al. this mechanism (Arnett et al., 2008; Baerwald & Barclay, 2011; Bennett & Hale, 2014; Horn et al., 2008), three of which suggest neutral behaviour towards wind turbine red aviation lighting, the fourth suggesting avoidance behaviour in one species (Lasiurus borealis) (Bennett & Hale, 2014). However, most of these studies were not primarily designed to assess bat responses to wind tur- bine red aviation lighting, and three of them were conducted only during the migration season (Baerwald & Barclay, 2011; Bennett & Hale, 2014; Horn et al., 2008). Furthermore, three of these stud- ies were based on bat fatalities (Arnett et al., 2008; Baerwald & Barclay, 2011; Bennett & Hale, 2014), addressing the direct con- sequences of collisions, but could not capture potential disruption of habitat use, such as the loss of habitat caused by avoidance behaviours (Barré et al., 2018; Ellerbrok et al., 2022; Gaultier et al., 2023; Minderman et al., 2017). Finally, these studies all in- volved both unlit wind turbines and turbines illuminated with red aviation lighting within a single wind farm, sometimes with both modalities located very close to each other, which could be prob- lematic given the potential spatial scale of the effect of aviation lighting on bats (Jonasson et al., 2024). In  this  study,  we  assessed  the  role  of  wind  turbine  lighting  in the behavioural responses (attraction and avoidance) of bats. To this end, we acoustically monitored bat activity at three wind farms in north- east Germany in June, outside both the peak mor- tality period and the migration season, although mortality events can still occur (Rydell et al., 2010). We used a sampling design based on triplets of sites simultaneously sampled: (i) at wind turbines illuminated throughout the night; (ii) at wind turbines illuminated only when an aircraft is detected with the ?Aircraft Detection Lighting System? (ADLS); and (iii) at control sites with- out wind  turbines within a  radius of 2?km. To consider potential  variations in bat responses to wind turbine lighting depending on the location relative to the turbine, we replicated this design at two different distances (at the base of the turbines in open habitats and at wooded edges located approximately 250?m from  the nearest wind turbine). We compared the responses of three functional bat guilds (short- range echolocators, SRE; medium- range  echolocators,  MRE;  long- range  echolocators,  LRE)  rep- resenting different levels of sensitivity to wind turbines (Barré et al., 2018; Roemer et al., 2019). We hypothesized that the activ- ity of bats may be different at wind turbines illuminated through- out the night compared to control sites, as a result of attraction or avoidance effects depending on the context (Barré et al., 2018; Ellerbrok et al., 2022, 2023; Leroux et al., 2022, 2024; Richardson et al., 2021; Sotillo et al., 2024). We also hypothesized that the use of the ADLS may mitigate the response of bats to wind tur- bines, thus potentially providing an easy- to- implement tool to reduce collision risks and habitat loss. We further expected differ- ences in activity between wind turbine lighting modalities to be more pronounced at the base of the turbines in open habitats for MRE and LRE, and to occur only at wooded edges located 250?m  from the nearest wind turbine for SRE. 2? |? MATERIAL S AND METHODS 2.1? |? Study area and wind farms sampled Data were collected at three wind farms in the Uckermark dis- trict of Brandenburg, in north- eastern Germany (53°23?6.4??N,  13°54?22.3??E)  (Figure 1). The land cover of the Uckermark district is dominated by crops, covering 65% of its area, in particular cereals, rapeseed and soya beans. This area was selected because it encom- passes one of the few wind farms to be equipped with the ADLS to date. The ADLS- equipped wind farm (?Kleisthöhe?, n?=?15  turbines)  is  located 11 and 22?km, respectively, from the two other wind farms.  All its turbines are equipped with red flashing lights (flash pattern: 1?s  ON, 0.5?s OFF, 1?s ON, 1.5?s OFF), as in turbines illuminated through- out the night, but these lighting systems are activated only when an aircraft is detected. Consequently, turbines at this wind farm were illuminated on average 12%?±?4% of the time at night, with no  consistent pattern of activation across nights (Appendix 1 in the Supporting Information). Aircraft were detected by the ADLS using RADAR at this wind farm. The other two sampled wind farms (?Gollmitz- Schönermark? and ?Grünberg?, with n?=?30  and  n?=?14  turbines,  respectively)  are  equipped with lighting systems that remain continuously activated throughout the night. These wind farms were selected for their sim- ilarities to the one equipped with the ADLS in terms of landscape context and turbine characteristics (Appendix 2 in the Supporting Information). Further details on the characteristics of the red avia- tion lighting at these three wind farms are provided in Appendix 3 in the Supporting Information. 2.2? |? Sampling design To assess whether wind turbine lighting may explain variations in bat activity levels near wind turbines, a triplet sampling design was conducted to simultaneously record bat activity at three sam- pling sites: (i) near a wind turbine illuminated throughout the night (?Constant WT site? hereafter), (ii) near a wind turbine partially illu- minated (?ADLS WT site? hereafter) and (iii) at a site without a wind turbine within a radius of 2?km (?control site? hereafter) (Figure 2). To ensure robust comparisons between wind turbine lighting mo- dalities, we selected sampling locations to minimize differences in wind turbine characteristics (hub height and rotor diameter) be- tween Constant WT sites and ADLS WT sites of the same triplet. Consequently, all sampled wind turbines had the same hub height (i.e.  100?m),  and  rotor  diameter  differed  by  less  than  3?m  on  av- erage within triplets (Appendix 4 in the Supporting Information). We also minimized variations in surrounding landscape features known to influence bat activity between Constant WT sites, ADLS WT sites and control sites of the same triplet (Appendix 5 in the Supporting Information). 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 4 of 13? |? ?? LARNOY et al. We replicated this design in two different habitats located at two distances from wind turbines (Figure 2). Bat activity was therefore recorded at nine triplets in open habitats at the base of the  turbines  (hereafter  ?OH?10?m?  dataset),  and  at  13  triplets  at  wooded edges (hedgerows or forest edges) located between 150 and 300?m from the nearest turbine (hereafter ?WH?250?m? dataset)  (average distance for Constant WT sites: 231.7?±?50.7?m; average  distance for ADLS WT sites: 243.6?±?49.2?m). For control sites, the  distance to the nearest wind turbine was 2564.8?±?133.1?m for the  WH?250?m dataset and 2498.3?±?543?m for the OH?10?m dataset.  Constant WT sites were sampled in both ?Gollmitz- Schönermark? and ?Grünberg? wind farms as part of both datasets (Gollmitz- Schönermark: n?=?4  for  the OH?10?m  dataset,  n?=?4  for  the WH? 250?m dataset; Grünberg: n?=?5 for the OH?10?m dataset, n?=?9 for  the WH?250?m dataset)  (Figure 1; Appendix 6 in the Supporting Information). This sampling design was motivated by four main considerations: (i) uncertainties regarding the spatial extent of the potential effect of wind turbine lighting on bats; (ii) the variability of their responses (i.e. attraction or avoidance) with habitat (Leroux et al., 2022); (iii) the limited availability of wooded edges at the base of wind tur- bines; and (iv) the need to ensure sufficient data for all bat guilds, F I G U R E 1?Location of the three wind farms sampled and sampling sites (Uckermark, Brandenburg, Germany). (a) Gollmitz- Schönermark  wind farm; (b) Kleisthöhe wind farm; (c) Grünberg wind farm. Open habitats: Sites in open habitats at the base of the turbines (OH?10?m);  Wooded edges: Sites at wooded edges (hedgerows or forest edges) located between 150 and 300?m from the nearest turbine (WH?250?m);  Constant: Sites near a wind turbine illuminated throughout the night (Constant WT sites); ADLS: Sites near a wind turbine partially illuminated (ADLS WT sites); control: Sites without a wind turbine within a radius of 2?km (control sites). Basemap: Google Satellite. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense ??? |? 5 of 13LARNOY et al. particularly SRE, which tend to avoid open habitats (Denzinger & Schnitzler, 2013; Frey- Ehrenbold et al., 2013). Sampling was conducted between 15 and 26 June 2021, with up to four triplets sampled the same night (Figure 1; Appendix 6 in the Supporting Information) for the OH?10?m dataset, and between  9 and 26 June 2021, with up to five triplets sampled the same night (Figure 1; Appendix 6 in the Supporting Information) for the WH? 250?m dataset. Each site was sampled over two consecutive nights.  For the WH?250?m dataset, only eight different turbines equipped  with the ADLS could be sampled due to a lack of hedgerows near other turbines. Consequently, five of these turbines were each in- volved in the sampling of two different triplets. Similarly, among the seven turbines illuminated throughout the night that were sampled for this dataset, three were sampled multiple times (two, three and four times). However, sites were consistently located at distinct hedgerows,  with  a minimum  distance  of  60?m  between  them.  No  ethical approval was required for data collection. When needed, fieldwork permissions were discussed and granted directly by the landowners. 2.3? |? Wind turbine operation Because wind turbine operation can influence bat activity, par- ticularly by reducing their activity at high rotation speeds (Cryan et al., 2014; Ellerbrok et al., 2024; Horn et al., 2008; Leroux et al., 2023), potential confounding effects with the tested wind turbine lighting modalities were also assessed (Appendix 7 in the Supporting Information). We found that they should not undermine the findings of this study (see Section 4). 2.4? |? Acoustic sampling We  recorded bat  echolocation  calls  using Song Meter SM4Bat FS  automatic passive acoustic recorders with omnidirectional SMM- U2  microphones  (Wildlife  Acoustics,  Inc.,  Concord, MA,  USA)  placed  1.5?m above the ground. All recorders were configured according to  the recommendations of the French bat monitoring program (FBMP)  (Millon et al., 2015)  and  recordings of all  sounds above 2?kHz  that  exceeded  the  background  noise  by  12?dB with  a  sampling  rate  of  384?kHz were carried out the entire night, from 30?min before sun- set until 30?min after sunrise and only under favourable conditions  following FBMP recommendations. We used the number of bat passes per night as a proxy of bat activity (e.g. Barré et al., 2018; Wickramasinghe et al., 2003). A bat pass was defined as the emission of one or more echoloca- tion calls by the same bat species during a 5- s interval. We used the TADARIDA software (Bas et al., 2017) to automatically detect echolocation calls and identify the taxon of each bat pass at the most accurate taxonomic level. A confidence score ranging from 0 to 1 was associated with each automatic identification. We F I G U R E 2?Schematic representation of the sampling design. OH?10?m: Sites in open habitats at the base of the turbines; WH?250?m:  Sites at wooded edges (hedgerows or forest edges) located between 150 and 300?m from the nearest turbine; Control site: Site without a  wind turbine within a radius of 2?km; ADLS: Site near a wind turbine partially illuminated; Constant aviation lighting: Site near a wind turbine  illuminated throughout the night. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 6 of 13? |? ?? LARNOY et al. conducted analyses only on bat passes with a confidence score superior to 0.5, and ensured the robustness of our findings by using a more conservative confidence score of 0.9. Indeed, each  threshold entails a different trade- off: applying a 0.5 threshold retains most true positives while discarding few false positives, whereas a 0.9 threshold removes most false positives but also a large number of true positives. Therefore, analysing data fil- tered using both thresholds (i.e. reflecting different trade- offs between false and true positives), and only drawing conclusive interpretations when both yield similar results, ensures the ro- bustness of our findings to automatic identification errors (Barré et al., 2019). Then, the identified taxa were grouped into three functional guilds based on the structure of their echolocation calls and, consequently,  their  foraging  strategies.  In  our  case,  the  SRE  in- cluded Myotis spp., Plecotus spp., Rhinolophus spp. and Barbastella barbastellus;  the  MRE  included  Pipistrellus spp., Hypsugo savii and Miniopterus schreibersii; and the LRE included Nyctalus spp., Eptesicus spp. and Tadarida teniotis (Denzinger & Schnitzler, 2013). These guilds feature different detection distances and different levels  of  sensitivity  to  wind  turbines.  Indeed,  the  SRE  are  par- ticularly vulnerable to habitat loss caused by wind turbines, fol- lowed by the MRE (Barré et al., 2018), whereas the LRE are highly sensitive  to  collision  risks,  also  followed  by  the  MRE  (Roemer  et al., 2019). This approach also enabled us to include species that are rare or difficult to detect?and therefore difficult to study indi- vidually?and to overcome most automatic identification errors by grouping species with similar echolocation calls that can be easily confused (Barré et al., 2019). We confirmed the absence of difference in identification errors between the different wind turbine lighting modalities by manually checking recordings using a stratified sampling (Appendix 8 in the Supporting Information). Indeed, the presence of a nearby wind tur- bine at Constant WT sites and ADLS WT sites may have led bats to modify their echolocation calls?particularly by increasing their frequency bandwidth?to better avoid this obstacle. These alter- ations in call characteristics may increase the acoustic overlap be- tween species, thereby raising the proportion of both false positives and false negatives for these two wind turbine lighting modalities. Differences in bat behaviour between Constant WT and ADLS WT sites (e.g. stronger attraction to Constant WT sites) could also have resulted in differences in identification error rates between these two wind turbine lighting modalities due to the varying proximity of bats to the wind turbines. 2.5? |? Statistical analysis To evaluate the effect of wind turbine lighting modality on bat activ- ity, we fitted generalized linear mixed models (GLMMs) with the glm- mTMB package (Brooks et al., 2017). Bat activity was included as the response variable (Equation 1). We used a negative binomial error distribution with a logarithmic link function to account for overdis- persion. Diagnostics of residuals were used to select between type I  and type II negative binomial models. We included the wind turbine  lighting modality?our variable of interest?as a three- level fixed ef- fect variable: ADLS WT site, Constant WT site and control site. We also added as a fixed effect the density of hedgerows computed in a buffer zone around the sampling site (expressed in linear length of hedgerows per surface area in km.km?2) to correct for residual but significant differences in landscape composition between sites within the same triplet (Appendix 5 in the Supporting Information). Finally, since our sampling design was based on triplets of sites sam- pled on the same nights we included the triplet as a random inter- cept effect. where Bat activityi,j is the number of bat passes observed at site i (of wind turbine lighting modality m and triplet l) on night j, and k is the dispersion parameter of the negative binomial. We fitted a total of 36 models, covering every combination of bat guild (SRE, MRE and LRE), dataset (OH?10?m and WH?250?m), auto- matic identification confidence score (0.5 and 0.9) and buffer size for computing hedgerow density (200, 500 and 1000?m, see Appendix 5 in the Supporting  Information). However, we reported in the main text only the results from the six models based on a confidence score of 0.5 and a buffer  size of 200?m, as using a 0.9 confidence  score is less conservative and resulted in a more restricted dataset, while a smaller buffer resulted in a lower overlap between sites. The other models were used to assess the sensitivity of our findings to variations in these two parameters and were reported in Appendix 9 in the Supporting Information. We checked that all explicative variables had a Variance Inflation  Factor inferior to 3.5, indicating the absence of multicollinearity (Zuur et al., 2010). We also assessed the quality of fit of models by checking the uniformity of the residual distribution, the homoge- neity of variance and the independence of the residuals using the DHARMA package (Hartig, 2022). Likelihood ratio tests (LRTs) (Fisher, 1922) were used to assess the effect of the wind turbine lighting modality on bat activity. When significant, we performed multiple pairwise comparisons using Tukey's post hoc tests (Tukey, 1949) to identify which modali- ties were significantly different from each other. All statistical anal- yses and graphs were performed using R 4.4.1 (R Core Team, 2024). Statistical significance was set at p?<?0.05. (1)Bat activityi,j ? Negative Binomial ( ?i,j ,?i,j + ?i,j 2 k ) log ( ?i,j ) =?+ M ? m ( ?m×Lightingmodalityi ) +?×Hedgerow densityi+?l+?i,j ?l ? N(0, ?) ?i,j ? N ( 0, ?? ) 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense ??? |? 7 of 13LARNOY et al. 3? |? RESULTS We recorded a total of 28,966 bat passes with a confidence score superior to 0.5 (see Appendix 10 in the Supporting  Information for  taxonomic  composition).  The  MRE  guild  was  the  most  abun- dant  (OH?10?m dataset: 83.9% of bat passes, n?=?4659; WH?250?m  dataset: 74.7%, n?=?17,492)  and  the  SRE  the  least  (OH?10?m  data- set: 2.1% of bat passes, n?=?115; WH?250?m dataset: 7.2%, n?=?1691)  (Appendix 10 in the Supporting  Information). All guilds were ob- served at all sampling sites for both datasets, except the SRE in the OH?10?m dataset (occurrence at 77.8% of sites). We found a significant effect of wind turbine lighting modality on bat activity in four out of six cases (Figure 3; Table 1). The activity of the SRE guild in the WH?250?m dataset, as well as that of the MRE  and LRE guilds  in  the OH?10?m dataset, was significantly higher at  sites with wind turbines illuminated throughout the night (?Constant WT? modality) than at sites without wind turbines (?control? modality) and at sites with turbines equipped with the ADLS (?ADLS WT? mo- dality), with no significant difference in activity between control and ADLS WT sites (Figure 3; Appendix 11 in the Supporting Information). A similar but nearly  significant pattern was observed  for  the MRE  guild in the WH?250?m dataset and SRE guild in the OH?10?m data- set (Figure 3; Appendix 11 in the Supporting Information). Similarly, in  the  WH?250?m  dataset,  the  activity  of  LRE  was  significantly  higher at Constant WT sites compared to ADLS WT sites (Figure 3; Appendix 11 in the Supporting Information). However, we observed no significant difference between Constant WT and control sites in this case and higher activity at control sites compared to ADLS WT sites (Figure 3; Appendix 11 in the Supporting Information). Finally, we found no significant effect of hedgerow density within a 200?m  radius on bat activity, except for  the MRE  in the OH?10?m dataset  where the effect was significantly negative (Table 1).  Models  in- cluding hedgerow density at 500, 1000?m or integrating bat passes  with a confidence score superior to 0.9 revealed similar patterns despite slight variations in significance (Appendix 9 in the Supporting Information). 4? |? DISCUSSION The role of red aviation lighting in bat responses to wind turbines has so far received little attention. This study provides the first empirical evidence that red aviation lighting drives at least partially attraction behaviour of all bat guilds towards wind turbines. Therefore, our re- sults demonstrate that the ADLS can cost- effectively contribute to mitigating the negative effects of wind turbine red aviation lighting on bats. 4.1? |? Bat responses to the ADLS Several bat genera (Rhinolophus, Hypsugo, Miniopterus and Tadarida) were automatically identified by the TADARIDA software, although  they do not occur in our study area. However, these identification errors were very unlikely to bias our findings as the correct guild was not identified for only 16 of these bat passes at a confidence score superior to 0.9: identifications of the Tadarida genus (n?=?14)  and Rhinolophus genus (n?=?2)  (Appendix 10 in the Supporting Information), that were actually non- bat noises (sounds emitted by small mammals for Rhinolophus). For all functional guilds, we observed that the acoustic activ- ity  under  and  250?m  from  wind  turbines  illuminated  throughout  the night was different (three cases out of six), or tended to be dif- ferent (two cases out of six), than at sites without a wind turbine within a radius of 2?km, in line with recent studies (Barré et al., 2018; Ellerbrok et al., 2022, 2023; Leroux et al., 2022, 2024; Richardson et al., 2021; Sotillo et al., 2024). The only exception was the LRE guild at 250?m from wind turbines, which exhibited a distinct trend.  Surprisingly, in all other cases, the direction of the difference was consistent, with greater bat activity near turbines. This indicates a local attraction behaviour towards wind turbines that might, in turn, increase collision risks. This consistency in bat responses to wind turbines was unanticipated. Indeed, the nature of this response (i.e.  attraction or avoidance) has been shown to depend on the distance from the turbines (Gaultier et al., 2023; Leroux et al., 2023), the local habitat (Leroux et al., 2022; Reusch et al., 2022; Scholz et al., 2025; Sotillo et al., 2024) and the species (Ellerbrok et al., 2024; Leroux et al., 2023; McKay  et  al., 2024). However, this result is in accor- dance with attraction behaviours that have mostly been reported at relatively small spatial scales, from below the turbine mast in a landscape context similar to that of this study (Leroux et al., 2022; Richardson et al., 2021),  to  400?m  in  forests  (Ellerbrok  et al., 2022). Our  findings  suggest  that  the  red  aviation  lighting may  be  an  important driving factor of bat attraction behaviour towards wind turbines.  Indeed, the activity at sites near wind turbines equipped  with ADLS was overall lower (four cases out of six), or tended to be lower (one case out of six), than at sites near wind turbines il- luminated throughout the night. Activity at these sites was also overall similar to the activity at control sites without wind turbines (five cases out of six). This difference in activity between wind tur- bines illuminated throughout the night and those equipped with ADLS cannot be attributed to the sometimes greater operation of the former (Appendix 7 in the Supporting Information). Such differ- ences may, in fact, have resulted in an underestimation of bat ac- tivity at sites near wind turbines illuminated throughout the night (Cryan et al., 2014; Ellerbrok et al., 2024; Horn et al., 2008; Leroux et al., 2023), leading to an underestimation of the difference in activ- ity between these two wind turbine lighting modalities. Nor can this  difference be attributed to the absence of infrared lights on wind turbines illuminated throughout the night sampled at the ?Grünberg? wind farm, unlike those equipped with ADLS (Appendix 3 in the Supporting Information). This may have indirectly attracted bats, as certain studies suggest that some insect species may respond to in- frared light (Callahan, 1965; Takács et al., 2009), thereby leading to an overestimation of bat activity at sites near wind turbines partially 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 8 of 13? |? ?? LARNOY et al. illuminated. Implementing ADLS could therefore contribute to main- taining activity levels near wind turbines similar to those observed without wind turbines nearby. To date, only four studies with contrasting results and conducted in North America on different bat communities have assessed the  effect of wind turbine red aviation lighting on bats, most of which F I G U R E 3?Estimated marginal mean of acoustic activity (in number of bat passes per night with an automatic identification confidence  score superior to 0.5) per bat guild (LRE, long- range echolocators; MRE, medium- range echolocators; SRE, short- range echolocators),  recorded at wooded edges (hedgerows or forest edges) located between 150 and 300?m from the nearest turbine (WH?250?m dataset, left  in green) or in open habitats at the base of the turbines (OH?10?m dataset, right in yellow) under the three wind turbine lighting modalities  (Constant WT, near wind turbines illuminated throughout the night; ADLS WT, near wind turbines equipped with aircraft detection lighting system; control, without wind turbines within a radius of 2?km) and with a radius buffer of 200?m for hedgerow density. Acoustic activity  indices with different letters are significantly different (p?<?0.05) according to pairwise comparisons (Tukey's method). The vertical bars  correspond to the 95% confidence interval. See Appendix 9 in the Supporting Information for results of models fitted with different radius buffers for hedgerow density (i.e. 500 and 1000?m) and acoustic data based on a confidence score superior to 0.9. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense ??? |? 9 of 13LARNOY et al. were  conducted  during  the migration  period.  Of  the  three  based  on bat fatalities, two did not report any red aviation lighting effect (Arnett et al., 2008; Baerwald & Barclay, 2011), in contrast to our study that suggests that bat attraction behaviour towards wind turbines could be caused by red aviation lighting. However, these studies were not primarily designed to assess this effect. Our find- ing also differs from the third study (Bennett & Hale, 2014), which reported avoidance behaviour towards illuminated wind turbines in one species (L. borealis) and neutral behaviour for the other species. However, it was carried out in a single wind farm with unlit turbines and turbines illuminated with red aviation lighting sometimes lo- cated very close together. Finally, Horn et al. (2008) observed a non- significant attraction behaviour towards red aviation lighting using thermal cameras. Besides, previous research has also assessed the effects of red artificial light on bats in a different context from wind energy (e.g. Barré et al., 2021, 2023; Spoelstra et al., 2017; Voigt et al., 2018). However, these findings are difficult to compare with our own, as the existing literature on this topic offers contradictory findings and wind turbine red aviation lighting is very specific (i.e. light flashes of very high intensity at height). Moreover, few of these  studies were conducted in early summer (but see Barré et al., 2021). Therefore, we consider that our research makes a major contribu- tion to understanding the role of wind turbines lighting in the be- havioural responses of bats. Another interesting result is the response pattern of LRE at wooded  edges  250?m  from  the wind  turbine,  a  guild  that  includes  the species most affected by mortality at wind turbines in Germany (Nyctalus noctula) (Rydell et al., 2010).  In this case, activity was still  higher at sites near wind turbines illuminated throughout the night than at sites near partially illuminated wind turbines, as in most other cases, suggesting that wind turbine red aviation lighting drives attraction behaviour. However, in this case, the activity levels at control sites were similar to those at sites near wind turbines illu- minated throughout the night. Here, this pattern could suggest the co- occurrence of two antagonistic responses of similar intensity in bats to wind turbines illuminated throughout the night. The first is an attraction behaviour driven by aviation lighting, as described above. The second is an avoidance behaviour, likely driven by another mechanism, that is revealed in the absence of attraction and results in lower activity at sites near partially illuminated wind turbines com- pared to control sites. This attraction behaviour towards wind turbine red aviation lighting could be direct, or indirect, due to the higher density of insects (Horn et al., 2008; McKay et al., 2024; Voigt, 2021). On the other hand, this avoidance behaviour could be related to the  operation of wind turbines and the resulting wake effect (Leroux et al., 2024), although this has not yet been demonstrated for this guild. 4.2? |? Implications for conservation This study revealed that smart lighting such as the ADLS can miti- gate the disruption of bat habitat use caused by wind turbines and associated collision risks. Furthermore, the use of the ADLS may also benefit other taxa that are strongly affected by collisions and for which aviation lighting could be a contributing factor, such as insects (Horn et al., 2008; Voigt, 2021) and migratory birds (Rebke et al., 2019). The ADLS also offers numerous benefits to wind turbine op- erators. First, this system is relatively easy- to- implement, as it re- quires no major modifications to existing wind turbines and several certified systems are already commercially available. Furthermore, the emission of light flashes over a reduced part of the night (ap- proximately 12% of the night in this study) and the resulting re- duction  in the ALAN generated by wind farms  is  likely to  increase  their acceptability to neighbouring populations. Last but not least, this potential mitigation measure does not affect the electricity pro- duction of wind turbines, unlike curtailment using blade feathering, which is currently the most common mitigation measure applied at wind farms. Consequently, given the ability of ADLS to reduce bat attraction to wind turbines and thus potentially lower bat fatalities, its implementation could represent a win- win conservation measure by avoiding a trade- off between energy production and biodiver- sity conservation.  In  light of all  these considerations, we advocate  TA B L E 1?Results of likelihood ratio tests (LRT) applied to generalized linear mixed models with a buffer of 200?m for the hedgerow  density variable (linear length per surface area in km.km?2), for each bat guild (LRE, long- range echolocators; MRE, medium- range  echolocators; SRE, short- range echolocators) and for bat passes with an automatic identification confidence score superior to 0.5 recorded at wooded edges (hedgerows or forest edges) located between 150 and 300?m from the nearest turbine (WH?250?m dataset) or in open  habitats at the base of the turbines (OH?10?m dataset). Guild Explanatory variable WH?250?m OH?10?m LRT p LRT p SRE Wind turbine lighting modality 33.04 <0.001 5.74 0.06 Hedgerow density (200?m) 0.07 0.79 0.39 0.53 MRE Wind turbine lighting modality 5.03 0.08 23.43 <0.001 Hedgerow density (200?m) 0.07 0.80 8.40 <0.01 LRE Wind turbine lighting modality 14.04 <0.001 9.40 <0.01 Hedgerow density (200?m) 3.21 0.07 0.41 0.52 Note: Significant p- values are shown in bold. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense 10 of 13? |? ?? LARNOY et al. for the wider deployment of the ADLS in combination with other existing mitigation measures. The mandatory use of the ADLS in Germany now may set a precedent encouraging broader interna- tional adoption. 4.3? |? Perspectives Although promising, our findings should be supplemented by similar studies conducted at different periods of the year as previous research has shown that bat responses to wind turbines may vary between sea- sons (Ellerbrok et al., 2022; McKay et al., 2024). Studies similar to this one should also be conducted at wind farms with a distinct surrounding landscape (e.g. mostly open fields or forest), as bat responses to wind turbines may also vary between habitats (Reusch et al., 2022; Sotillo et al., 2024). Such complementary studies would allow generalizing the ability of ADLS to mitigate bat attraction behaviour towards wind turbines throughout the year and across different landscapes. Although our results allow drawing conclusions about the po- tential of ADLS to reduce bat attraction towards wind turbines, we cannot conclude with certainty about its potential to mitigate bat fatalities since we assessed acoustic activity at ground height and outside the peak mortality period (Rydell et al., 2010). Further re- search conducted in late summer and monitoring acoustic activity at wind turbine nacelles or, better still, bat carcasses around turbines in late summer, would enable that assertion to be validated. Finally, the precise mechanism underlying bat attraction towards wind turbine red aviation lighting remains to be elucidated. Future studies with a similar sampling design could compare foraging ac- tivity between the different lighting modalities. Such analyses might help to discriminate between direct and indirect attraction of bats towards wind turbines. Similarly, assessing the relationship between acoustic bat activity at wind turbines with ADLS and the time of ac- tivation of the system during the night, although it requires a larger dataset, could provide a better understanding of the time scale of this mechanism. AUTHOR CONTRIBUTIONS Christian Kerbiriou, Isabelle Le Viol, Pauline Lefebvre, Nicolas Valet,  Kévin Barré and Camille Leroux conceived the ideas and designed the  methodology;  Gaëlle  Larnoy,  Pauline  Lefebvre  and  Camille  Leroux collected the data; Gaëlle Larnoy and Fabien Verniest led the analysis of the data with the support of Christian Kerbiriou, Isabelle Le Viol, Kévin Barré and Camille Leroux; Gaëlle Larnoy and  Fabien Verniest led the writing of the manuscript with the support of Christian Kerbiriou, Isabelle Le Viol, Pauline Lefebvre, Nicolas Valet,  Kévin Barré and Camille Leroux. All authors contributed critically to the drafts and gave final approval for publication. ACKNOWLEDG EMENTS We thank two anonymous reviewers for comments that significantly improved the quality of the manuscript. We are grateful to C. Herminet for her contribution to the study design and data collection. We also thank E. Trébuchet and M. Clément- Lacroix for the manual  verification of automatic identifications, Q. Grisouard for helping with the figures and P. Bach and C. Roemer for helping with the German  version of the abstract. We are grateful to Dark Sky, ENERTRAG SE,  ENERTRAG  Systemtechnik,  the  Kompetenzzentrum  Naturschutz  und Energiewende, the Landesamt für Umwelt Brandenburg, the Ministerium  für  Infrastruktur  und  Landesplanung,  the  Uckermark  district, M. Fritze, Y. Gager and C. Voigt for their invaluable help in the  search for information on the operation and characteristics of wind turbines. This work was supported by the Agence de la transition écologique (ADEME), the Association Nationale de la Recherche et  de la Technologie (Grant No. 2019/1566) and Auddicé biodiversité. CONFLIC T OF INTERE S T S TATEMENT This work was  initiated  as  part  of Camille  Leroux's  PhD  research,  which was co- supervised by Christian Kerbiriou, Isabelle Le Viol and  Kévin Barré from the National Museum of Natural History (MNHN)  and Nicolas Valet from Auddicé biodiversité. Auddicé biodiversité is  an environmental consultancy that conducts wind farm impact as- sessment studies. At the time of submission, one of the authors? Camille Leroux?was working at Auddicé biodiversité. This work continued  as  part  of  Gaëlle  Larnoy's  Master's  thesis  and  Fabien  Verniest's postdoctoral position, which were mainly funded by ADEME, a public agency promoting renewable energies. Members  of the wind energy sector financed part of the bat recorders and provided some technical data and expertise on wind turbine opera- tion and features, as stated above. Thus, the authors declare a po- tential conflict of interest. However, sampling design, acoustic data collection, analysis and writing were conducted only by the authors and members of the wind energy sector did not contribute to the draft. Furthermore, sampling design and sampling sites were de- termined independently from Auddicé biodiversité activities, and identification of bat echolocation calls and bat activity measures were  provided  by  TADARIDA  software,  a MNHN web  portal,  ex- cept for manual verifications of automatic identification errors that were  performed  by  Elise  Trébuchet  and Margot  Clément- Lacroix,  who were working at Auddicé biodiversité at the time of submis- sion. The authors certify that the collaboration did not interfere with the stated hypothesis, the way it was tested or the interpretations and conclusions. Authors take full responsibility for the integrity of the study. DATA AVAIL ABILIT Y S TATEMENT Data available from the Zenodo open repository https:// doi. org/ 10. 5281/ zenodo. 17454526 (Larnoy et al., 2025). ORCID Fabien Verniest https://orcid.org/0000-0001-5744-3185 Christian Kerbiriou https://orcid.org/0000-0001-6080-4762 Isabelle Le Viol https://orcid.org/0000-0003-3475-5615 Kévin Barré https://orcid.org/0000-0001-5368-4053 Camille Leroux https://orcid.org/0000-0002-4984-3485 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.5281/zenodo.17454526 https://doi.org/10.5281/zenodo.17454526 https://orcid.org/0000-0001-5744-3185 https://orcid.org/0000-0001-5744-3185 https://orcid.org/0000-0001-6080-4762 https://orcid.org/0000-0001-6080-4762 https://orcid.org/0000-0003-3475-5615 https://orcid.org/0000-0003-3475-5615 https://orcid.org/0000-0001-5368-4053 https://orcid.org/0000-0001-5368-4053 https://orcid.org/0000-0002-4984-3485 https://orcid.org/0000-0002-4984-3485 ??? |? 11 of 13LARNOY et al. R E FE R E N C E S Adams, E. M., Gulka, J., & Williams, K. A. (2021). A review of the effec- tiveness of operational curtailment for reducing bat fatalities at ter- restrial wind farms in North America. PLoS One, 16(11), e0256382. https:// doi. org/ 10. 1371/ journ al. pone. 0256382 Arnett, E. B., Brown, W. K., Erickson, W. P., Fiedler, J. K., Hamilton, B. L.,  Henry, T. H., & Tankersley, R. D.  (2008). Patterns of bat  fatalities  at wind energy  facilities  in North America. The Journal of Wildlife Management, 72(1), 61?78. Baerwald, E. F., & Barclay, R. M. R. (2011). Patterns of activity and fatality  of migratory bats at a wind energy facility in Alberta, Canada. The Journal of Wildlife Management, 75(5), 1103?1114. https:// doi. org/ 10. 1002/ jwmg. 147 Barré, K., Froidevaux, J. S. P., Leroux, C., Mariton, L., Fritze, M., Kerbiriou,  C.,  &  Roemer,  C.  (2022).  Over  a  decade  of  failure  to  implement  UNEP/EUROBATS  guidelines  in wind  energy  planning:  A  call  for  action. Conservation Science and Practice, 4(11), e12805. https:// doi. org/ 10. 1111/ csp2. 12805 Barré, K., Kerbiriou, C., Ing, R.- K., Bas, Y., Azam, C., Le Viol, I., & Spoelstra,  K. (2021). Bats seek refuge in cluttered environment when exposed to white and red lights at night. Movement Ecology, 9(1), 3. https:// doi. org/ 10. 1186/ s4046 2- 020- 00238 - 2 Barré, K., Le Viol, I., Bas, Y., Julliard, R., & Kerbiriou, C. (2018). Estimating  habitat loss due to wind turbine avoidance by bats: Implications for  European siting guidance. Biological Conservation, 226,  205?214.  https:// doi. org/ 10. 1016/j. biocon. 2018. 07. 011 Barré, K., Le Viol, I., Julliard, R., Pauwels, J., Newson, S. E., Julien, J.- F.,  &  Bas,  Y.  (2019).  Accounting  for  automated  identification  errors  in acoustic surveys. Methods in Ecology and Evolution, 10(8), 1171? 1188. https:// doi. org/ 10. 1111/ 2041-  210X. 13198  Barré,  K.,  Thomas,  I.,  Le  Viol,  I.,  Spoelstra,  K.,  &  Kerbiriou,  C.  (2023).  Manipulating spectra of artificial  light affects movement patterns  of bats along ecological corridors. Animal Conservation, 26(6), 865? 875. https:// doi. org/ 10. 1111/ acv. 12875 Bas, Y., Bas, D., & Julien, J.- F. (2017). Tadarida: A toolbox for animal de- tection on acoustic recordings. Journal of Open Research Software, 5(1), 6. https:// doi. org/ 10. 5334/ jors. 154 Bennett, V. J., & Hale, A. M. (2014). Red aviation lights on wind turbines  do not  increase bat?turbine collisions. Animal Conservation, 17(4), 354?358. https:// doi. org/ 10. 1111/ acv. 12102 Brooks, M. E., Kristensen, K., Benthem, K., J., van Magnusson, A., Berg,  C. W., Nielsen, A., Skaug, H. J., Mächler, M., & Bolker, B. M. (2017).  glmmTMB balances speed and flexibility among packages for zero-  inflated generalized linear mixed modeling. The R Journal, 9(2), 378. https:// doi. org/ 10. 32614/ RJ- 2017- 066 Callahan, P. S. (1965). Far infra- red emission and detection by night- flying  moths. Nature, 206(4989),  1172?1173.  https:// doi. org/ 10. 1038/ 20611 72a0 Cryan, P. M. (2008). Mating behavior as a possible cause of bat fatalities  at wind turbines. The Journal of Wildlife Management, 72(3), 845? 849. https:// doi. org/ 10. 2193/ 2007- 371 Cryan,  P.  M.,  &  Barclay,  R.  M.  R.  (2009).  Causes  of  bat  fatalities  at wind turbines: Hypotheses and predictions. Journal of Mammalogy, 90(6),  1330?1340.  https:// doi. org/ 10. 1644/ 09- MAMM-  S-  076R1. 1 Cryan, P. M., Gorresen, P. M., Hein, C. D., Schirmacher, M. R., Diehl, R.  H.,  Huso, M. M.,  Hayman, D.  T.,  Fricker,  P.  D.,  Bonaccorso,  F.  J.,  Johnson, D. H., Heist, K., & Dalton, D. C. (2014). Behavior of bats at wind turbines. Proceedings of the National Academy of Sciences of the United States of America, 111(42), 15126?15131. https:// doi. org/ 10. 1073/ pnas. 14066 72111 Denzinger, A., & Schnitzler, H.- U. (2013). Bat guilds, a concept to classify the highly diverse foraging and echolocation behaviors of micro- chiropteran bats. Frontiers in Physiology, 4, 164. https:// doi. org/ 10. 3389/ fphys. 2013. 00164 Duriez,  O.,  Pilard,  P.,  Saulnier,  N.,  Boudarel,  P.,  &  Besnard,  A.  (2023).  Windfarm collisions in medium- sized raptors: Even increasing popu- lations can suffer strong demographic impacts. Animal Conservation, 26(2), 264?275. https:// doi. org/ 10. 1111/ acv. 12818 Ellerbrok,  J.  S.,  Delius,  A.,  Peter,  F.,  Farwig,  N.,  &  Voigt,  C.  C.  (2022).  Activity of forest specialist bats decreases towards wind turbines at forest sites. Journal of Applied Ecology, 59(10), 2497?2506. https:// doi. org/ 10. 1111/ 1365- 2664. 14249 Ellerbrok,  J.  S.,  Farwig, N.,  Peter,  F.,  Rehling,  F., & Voigt, C. C.  (2023).  Forest gaps around wind turbines attract bat species with high col- lision risk. Biological Conservation, 288, 110347. https:// doi. org/ 10. 1016/j. biocon. 2023. 110347 Ellerbrok, J. S., Farwig, N., Peter, F., & Voigt, C. C. (2024). Forest bat ac- tivity declines with increasing wind speed in proximity of operating wind turbines. Global Ecology and Conservation, 49, e02782. https:// doi. org/ 10. 1016/j. gecco. 2023. e02782 Fisher,  R.  A.  (1922).  On  the  mathematical  foundations  of  theoretical  statistics. Philosophical Transactions of the Royal Society of London. Series A, Mathematical and Physical Sciences, 222(594?604),  309? 368. https:// doi. org/ 10. 1098/ rsta. 1922. 0009 Frey- Ehrenbold,  A.,  Bontadina,  F.,  Arlettaz,  R.,  &  Obrist,  M.  K.  (2013).  Landscape connectivity, habitat structure and activity of bat guilds in farmland- dominated matrices. Journal of Applied Ecology, 50(1), 252?261. https:// doi. org/ 10. 1111/ 1365- 2664. 12034 Frick, W. F., Baerwald, E. F., Pollock, J. F., Barclay, R. M. R., Szymanski, J. A.,  Weller, T.  J., & McGuire,  L. P.  (2017). Fatalities at wind  turbines may  threaten population viability of a migratory bat. Biological Conservation, 209, 172?177. https:// doi. org/ 10. 1016/j. biocon. 2017. 02. 023 Gaultier, S. P., Lilley, T. M., Vesterinen, E. J., & Brommer, J. E. (2023). The  presence of wind turbines repels bats in boreal forests. Landscape and Urban Planning, 231, 104636. https:// doi. org/ 10. 1016/j. landu rbplan. 2022. 104636 Guest, E. E., Stamps, B. F., Durish, N. D., Hale, A. M., Hein, C. D., Morton,  B. P., & Fritts, S. R.  (2022). An updated review of hypotheses  re- garding bat attraction to wind turbines. Animals, 12(3), 343. https:// doi. org/ 10. 3390/ ani12 030343 Hartig, F. (2022). DHARMa: Residual diagnostics for hierarchical (multi- level/mixed) regression models. R package version 0.4.5. https:// CRAN. R-  proje ct. org/ packa ge= DHARMa Horn, J. W., Arnett, E. B., & Kunz, T. H. (2008). Behavioral responses of bats to operating wind turbines. The Journal of Wildlife Management, 72(1), 123?132. https:// doi. org/ 10. 2193/ 2006- 465 Hutchinson, M., & Zhao, F. (2023). Global wind report 2023. Global wind energy council, 3. IPCC.  (2022). Climate change 2022: Mitigation of climate change.  In P.  R. Shukla, J. Skea, R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum, & J. Malley (Eds.), Contribution of working group III to the sixth assessment report of the Intergovernmental Panel on Climate Change.  Cambridge  University  Press.  https:// doi. org/ 10. 1017/ 97810 09157 926. 006 IRENA.  (2022).  Renewable energy statistics 2022.  The  International  Renewable Energy Agency. Jonasson, K. A., Adams, A. M., Brokaw, A. F., Whitby, M. D., O'Mara, M.  T., & Frick, W. F. (2024). A multisensory approach to understand- ing bat responses to wind energy developments. Mammal Review, 54(3), 229?242. https:// doi. org/ 10. 1111/ mam. 12340 Kunz, T. H., Arnett, E. B., Erickson, W. P., Hoar, A. R., Johnson, G. D., Larkin,  R. P., & Tuttle, M. D. (2007). Ecological impacts of wind energy devel- opment on bats: Questions, research needs, and hypotheses. Frontiers in Ecology and the Environment, 5(6), 315?324. https:// doi. org/ 10. 1890/ 1540-  9295(2007) 5[315: EIOWED] 2.0. CO; 2 Kuvlesky, W. P., Brennan, L. A., Morrison, M. L., Boydston, K. K., Ballard,  B. M., & Bryant, F. C. (2007). Wind energy development and wildlife  conservation: Challenges and opportunities. The Journal of Wildlife Management, 71(8), 2487?2498. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.1371/journal.pone.0256382 https://doi.org/10.1002/jwmg.147 https://doi.org/10.1002/jwmg.147 https://doi.org/10.1111/csp2.12805 https://doi.org/10.1111/csp2.12805 https://doi.org/10.1186/s40462-020-00238-2 https://doi.org/10.1186/s40462-020-00238-2 https://doi.org/10.1016/j.biocon.2018.07.011 https://doi.org/10.1111/2041-210X.13198 https://doi.org/10.1111/acv.12875 https://doi.org/10.5334/jors.154 https://doi.org/10.1111/acv.12102 https://doi.org/10.32614/RJ-2017-066 https://doi.org/10.1038/2061172a0 https://doi.org/10.1038/2061172a0 https://doi.org/10.2193/2007-371 https://doi.org/10.1644/09-MAMM-S-076R1.1 https://doi.org/10.1644/09-MAMM-S-076R1.1 https://doi.org/10.1073/pnas.1406672111 https://doi.org/10.1073/pnas.1406672111 https://doi.org/10.3389/fphys.2013.00164 https://doi.org/10.3389/fphys.2013.00164 https://doi.org/10.1111/acv.12818 https://doi.org/10.1111/1365-2664.14249 https://doi.org/10.1111/1365-2664.14249 https://doi.org/10.1016/j.biocon.2023.110347 https://doi.org/10.1016/j.biocon.2023.110347 https://doi.org/10.1016/j.gecco.2023.e02782 https://doi.org/10.1016/j.gecco.2023.e02782 https://doi.org/10.1098/rsta.1922.0009 https://doi.org/10.1111/1365-2664.12034 https://doi.org/10.1016/j.biocon.2017.02.023 https://doi.org/10.1016/j.landurbplan.2022.104636 https://doi.org/10.1016/j.landurbplan.2022.104636 https://doi.org/10.3390/ani12030343 https://doi.org/10.3390/ani12030343 https://cran.r-project.org/package=DHARMa https://cran.r-project.org/package=DHARMa https://doi.org/10.2193/2006-465 https://doi.org/10.1017/9781009157926.006 https://doi.org/10.1017/9781009157926.006 https://doi.org/10.1111/mam.12340 https://doi.org/10.1890/1540-9295(2007)5%5B315:EIOWED%5D2.0.CO;2 https://doi.org/10.1890/1540-9295(2007)5%5B315:EIOWED%5D2.0.CO;2 12 of 13? |? ?? LARNOY et al. Larnoy, G., Verniest, F., Kerbiriou, C., Le Viol, I., Lefebvre, P., Valet, N.,  Barré, K., &  Leroux, C.  (2025). Dataset  for  ?Minimizing  aviation  lighting duration reduces bat attraction to wind turbines? [Data set]. Zenodo. https:// doi. org/ 10. 5281/ zenodo. 17454526 Leroux, C., Barré, K., Valet, N., Kerbiriou, C., & Le Viol, I. (2024). Distribution  of common pipistrelle (Pipistrellus pipistrellus) activity is altered by airflow disruption generated by wind turbines. PLoS One, 19(5), e0303368. https:// doi. org/ 10. 1371/ journ al. pone. 0303368 Leroux, C., Kerbiriou, C., Le Viol, I., Valet, N., & Barré, K. (2022). Distance  to hedgerows drives local repulsion and attraction of wind turbines on bats: Implications for spatial siting. Journal of Applied Ecology, 59(8), 2142?2153. https:// doi. org/ 10. 1111/ 1365- 2664. 14227 Leroux,  C.,  Le  Viol,  I.,  Valet,  N.,  Kerbiriou,  C.,  &  Barré,  K.  (2023).  Disentangling mechanisms responsible for wind energy effects on European bats. Journal of Environmental Management, 346, 118987. https:// doi. org/ 10. 1016/j. jenvm an. 2023. 118987 McKay, R. A.,  Johns, S. E., Bischof, R., Matthews, F., Van Der Kooij,  J.,  Yoh, N., & Eldegard, K. (2024). Wind energy development can lead  to guild- specific habitat loss in boreal forest bats. Wildlife Biology, 2024, e01168. https:// doi. org/ 10. 1002/ wlb3. 01168 Millon,  L.,  Julien,  J.- F.,  Julliard,  R.,  &  Kerbiriou,  C.  (2015).  Bat  activity  in intensively farmed landscapes with wind turbines and offset measures. Ecological Engineering, 75,  250?257. https:// doi. org/ 10. 1016/j. ecole ng. 2014. 11. 050 Minderman,  J.,  Gillis, M.,  Daly,  H.,  &  Park,  K.  (2017).  Landscape- scale  ef- fects of single- and multiple small wind turbines on bat activity. Animal Conservation, 20, 455?462. https:// doi. org/ 10. 1111/ acv. 12331 O'Shea, T. J., Cryan, P. M., Hayman, D. T. S., Plowright, R. K., & Streicker, D.  G. (2016). Multiple mortality events in bats: A global review. Mammal Review, 46(3), 175?190. https:// doi. org/ 10. 1111/ mam. 12064 R Core Team. (2024). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https:// www. R- proje ct. org/ Rebke, M., Dierschke, V., Weiner, C. N., Aumüller, R., Hill, K., & Hill, R.  (2019). Attraction of nocturnally migrating birds to artificial light: The influence of colour, intensity and blinking mode under differ- ent cloud cover conditions. Biological Conservation, 233, 220?227.  https:// doi. org/ 10. 1016/j. biocon. 2019. 02. 029 Reusch, C., Lozar, M., Kramer- Schadt, S., & Voigt, C. C.  (2022). Coastal  onshore wind turbines lead to habitat loss for bats in northern Germany. Journal of Environmental Management, 310, 114715. https:// doi. org/ 10. 1016/j. jenvm an. 2022. 114715 Richardson, S. M., Lintott, P. R., Hosken, D. J., Economou, T., & Mathews,  F.  (2021). Peaks  in bat activity at  turbines and  the  implications  for  mitigating the impact of wind energy developments on bats. Scientific Reports, 11(1), 3636. https:// doi. org/ 10. 1038/ s4159 8- 021- 82014 - 9 Rodrigues, L., Bach, L., Dubourg- Savage, M., Karapandza, B., Kovac, D.,  Kervyn, T., Dekker, J., Kepel, A., Bach, P., Collins, J., Harbusch, C.,  Park, K., Micevski, B., & Minderman,  J.  (2015). Guidelines for con- sideration of bats in wind farm projects - revision 2014. EUROBATS  Publication Series No. 6 (English Version). Roemer, C., Bas, Y., Disca, T., & Coulon, A. (2019). Influence of landscape  and time of year on bat- wind turbines collision risks. Landscape Ecology, 34(12),  2869?2881.  https:// doi. org/ 10. 1007/ s1098 0- 019- 00927 - 3 Rydell,  J., Bach, L., Dubourg- Savage, M.- J., Green, M., Rodrigues, L., &  Hedenström, A. (2010). Bat mortality at wind turbines in north- western Europe. Acta Chiropterologica, 12,  261?274.  https:// doi. org/ 10. 3161/ 15081 1010X 537846 Saidur, R., Rahim, N. A., Islam, M. R., & Solangi, K. H. (2011). Environmental  impact of wind energy. Renewable and Sustainable Energy Reviews, 15(5), 2423?2430. https:// doi. org/ 10. 1016/j. rser. 2011. 02. 024 Scholz, C., Klein, H., & Voigt, C. C. (2025). Wind turbines displace bats from drinking sites. Biological Conservation, 302, 110968. https:// doi. org/ 10. 1016/j. biocon. 2025. 110968 Schuster, E., Bulling, L., & Köppel, J. (2015). Consolidating the state of knowledge: A synoptical review of wind energy's wildlife effects. Environmental Management, 56(2),  300?331.  https:// doi. org/ 10. 1007/ s0026 7- 015- 0501- 5 Sotillo,  A.,  Le  Viol,  I.,  Barré,  K.,  Bas,  Y.,  &  Kerbiriou,  C.  (2024).  Context-  dependent effects of wind turbines on bats in rural landscapes. Biological Conservation, 295, 110647. https:// doi. org/ 10. 1016/j. biocon. 2024. 110647 Spoelstra, K., van Grunsven, R. H. A., Ramakers, J. J. C., Ferguson, K. B., Raap, T., Donners, M., & Visser, M. E.  (2017). Response of bats to  light with  different spectra: Light- shy and agile bat presence is affected by white and green, but not red light. Proceedings of the Royal Society B: Biological Sciences, 284(1855), 20170075. https:// doi. org/ 10. 1098/ rspb. 2017. 0075 Stone, E. L., Harris, S., & Jones, G. (2015). Impacts of artificial lighting on  bats: A review of challenges and solutions. Mammalian Biology, 80(3), 213?219. https:// doi. org/ 10. 1016/j. mambio. 2015. 02. 004 Takács, S., Bottomley, H., Andreller, I., Zaradnik, T., Schwarz, J., Bennett,  R., Strong, W., & Gries, G. (2009). Infrared radiation from hot cones  on cool conifers attracts seed- feeding insects. Proceedings of the Royal Society B: Biological Sciences, 276(1657), 649?655. https:// doi. org/ 10. 1098/ rspb. 2008. 0742 Tukey, J. W. (1949). Comparing individual means in the analysis of vari- ance. Biometrics, 5(2), 99. https:// doi. org/ 10. 2307/ 3001913 Voigt, C. C. (2021). Insect fatalities at wind turbines as biodiversity sinks.  Conservation Science and Practice, 3(5), e366. https:// doi. org/ 10. 1111/ csp2. 366 Voigt, C. C., Dekker,  J.,  Fritze, M., Gazaryan,  S., Hölker,  F.,  Jones, G., &  Zagmajster, M.  (2021). The  impact of  light pollution on bats varies  according to foraging guild and habitat context. Bioscience, 71(10), 1103?1109. https:// doi. org/ 10. 1093/ biosci/ biab087 Voigt, C. C., Rehnig, K., Lindecke, O., & Petersons, G. (2018). Migratory bats  are attracted by red light but not by warm- white light: Implications  for the protection of nocturnal migrants. Ecology and Evolution, 8(18), 9353?9361. https:// doi. org/ 10. 1002/ ece3. 4400 Voigt, C. C., Straka, T. M., & Fritze, M. (2019). Producing wind energy at  the cost of biodiversity: A stakeholder view on a green- green di- lemma. Journal of Renewable and Sustainable Energy, 11(6), 063303. https:// doi. org/ 10. 1063/1. 5118784 Wang, S., & Wang, S.  (2015).  Impacts of wind energy on environment:  A review. Renewable and Sustainable Energy Reviews, 49, 437?443.  https:// doi. org/ 10. 1016/j. rser. 2015. 04. 137 Whitby, M. D., O'Mara, M. T., Hein, C. D., Huso, M., & Frick, W. F. (2024).  A decade of curtailment studies demonstrates a consistent and ef- fective strategy to reduce bat fatalities at wind turbines in North  America. Ecological Solutions and Evidence, 5(3), e12371. https:// doi. org/ 10. 1002/ 2688- 8319. 12371 Wickramasinghe, L. P., Harris, S., Jones, G., & Vaughan, N. (2003). Bat activ- ity and species richness on organic and conventional farms: Impact of  agricultural intensification. Journal of Applied Ecology, 40(6), 984?993.  https:// doi. org/ 10. 1111/j. 1365- 2664. 2003. 00856. x Zuur, A. F.,  Ieno, E. N., & Elphick, C. S.  (2010). A protocol  for data ex- ploration to avoid common statistical problems: Data exploration. Methods in Ecology and Evolution, 1(1),  3?14.  https:// doi. org/ 10. 1111/j. 2041-  210X. 2009. 00001. x SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. Appendix S1. Duration of lighting activation for wind turbines equipped with ADLS. Appendix S2. Wind turbines characteristics of the three wind farm sampled. Appendix S3. Lighting characteristics of sampled wind turbines. Appendix S4. Assessment of confounding effects between lighting modality and wind turbine characteristics. 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.5281/zenodo.17454526 https://doi.org/10.1371/journal.pone.0303368 https://doi.org/10.1111/1365-2664.14227 https://doi.org/10.1016/j.jenvman.2023.118987 https://doi.org/10.1002/wlb3.01168 https://doi.org/10.1016/j.ecoleng.2014.11.050 https://doi.org/10.1016/j.ecoleng.2014.11.050 https://doi.org/10.1111/acv.12331 https://doi.org/10.1111/mam.12064 https://www.r-project.org/ https://doi.org/10.1016/j.biocon.2019.02.029 https://doi.org/10.1016/j.jenvman.2022.114715 https://doi.org/10.1038/s41598-021-82014-9 https://doi.org/10.1007/s10980-019-00927-3 https://doi.org/10.1007/s10980-019-00927-3 https://doi.org/10.3161/150811010X537846 https://doi.org/10.3161/150811010X537846 https://doi.org/10.1016/j.rser.2011.02.024 https://doi.org/10.1016/j.biocon.2025.110968 https://doi.org/10.1016/j.biocon.2025.110968 https://doi.org/10.1007/s00267-015-0501-5 https://doi.org/10.1007/s00267-015-0501-5 https://doi.org/10.1016/j.biocon.2024.110647 https://doi.org/10.1016/j.biocon.2024.110647 https://doi.org/10.1098/rspb.2017.0075 https://doi.org/10.1016/j.mambio.2015.02.004 https://doi.org/10.1098/rspb.2008.0742 https://doi.org/10.1098/rspb.2008.0742 https://doi.org/10.2307/3001913 https://doi.org/10.1111/csp2.366 https://doi.org/10.1111/csp2.366 https://doi.org/10.1093/biosci/biab087 https://doi.org/10.1002/ece3.4400 https://doi.org/10.1063/1.5118784 https://doi.org/10.1016/j.rser.2015.04.137 https://doi.org/10.1002/2688-8319.12371 https://doi.org/10.1002/2688-8319.12371 https://doi.org/10.1111/j.1365-2664.2003.00856.x https://doi.org/10.1111/j.2041-210X.2009.00001.x https://doi.org/10.1111/j.2041-210X.2009.00001.x ??? |? 13 of 13LARNOY et al. Appendix S5. Assessment of confounding effects between lighting modality and landscape. Appendix S6. Sampling dates of triplets. Appendix S7. Assessment of confounding effects between lighting modality and operation. Appendix S8. Verification of automatic identification. Appendix S9. Sensitivity of results to automatic identification confidence score and radius buffer for hedgerow density. Appendix S10. Number, percentage and occurrence of bat passes by  guild and taxon. Appendix S11. Results of pairwise comparison tests. How to cite this article: Larnoy, G., Verniest, F., Kerbiriou, C., Le Viol, I., Lefebvre, P., Valet, N., Barré, K., & Leroux, C.  (2026). Minimizing aviation lighting duration reduces bat  attraction to wind turbines. Journal of Applied Ecology, 63, e70226. https://doi.org/10.1111/1365-2664.70226 13652664, 2026, 1, D ow nloaded from https://besjournals.onlinelibrary.w iley.com /doi/10.1111/1365-2664.70226 by IN SE E , W iley O nline L ibrary on [16/01/2026]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://doi.org/10.1111/1365-2664.70226  Minimizing aviation lighting duration reduces bat attraction to wind turbines  Abstract  1 | INTRODUCTION  2 | MATERIALS AND METHODS  2.1 | Study area and wind farms sampled  2.2 | Sampling design  2.3 | Wind turbine operation  2.4 | Acoustic sampling  2.5 | Statistical analysis  3 | RESULTS  4 | DISCUSSION  4.1 | Bat responses to the ADLS  4.2 | Implications for conservation  4.3 | Perspectives  AUTHOR CONTRIBUTIONS  ACKNOWLEDGEMENTS  CONFLICT OF INTEREST STATEMENT  DATA AVAILABILITY STATEMENT  ORCID  REFERENCES