Managing hedgerows for biodiversity: Disentangling the effects of trimming, structure and connectivity on the use of linear features by bats
Data files
Oct 16, 2025 version files 704.69 KB
-
Dataset_hegderow_bat_activity_2025.csv
691.95 KB
-
README.md
12.74 KB
Oct 16, 2025 version files 704.69 KB
-
Dataset_hegderow_bat_activity_2025.csv
691.95 KB
-
README.md
12.74 KB
Abstract
Hedgerows are key semi-natural elements in European agricultural landscapes supporting diverse wildlife, including bats, that depend on these linear elements for foraging and commuting. Despite replanting initiatives, many hedgerows remain degraded due to intensive management practices such as over-trimming. To develop effective and generalisable conservation guidelines, more research is needed to understand how both management practices and intrinsic hedgerow characteristics influence bat species with different foraging strategies.
We conducted passive acoustic monitoring across 448 hedgerows to assess the influence of six characteristics on bat activity: trimming method, height, width, foliage density, connectivity, and woody plant diversity. We examined responses for two species associated with hedgerow landscapes (Barbastella barbastellus, Rhinolophus ferrumequinum), and three echolocation guilds: short-, mid-, and long-range echolocators (SRE, MRE, LRE), corresponding to clutter, edge, and open space foragers, respectively.
Pollarding consistently increased bat activity, especially for tree-dwelling species, likely by maintaining mature trees with larger crowns and diversified microhabitats. Coppicing generally reduced activity for SRE, B. barbastellus, and R. ferrumequinum, while increasing LRE activity. Taller, tree-filled hedgerows generally supported higher bat activity, whereas width and foliage density had more limited effects. Finally, uninterrupted hedgerow segments enhanced SRE and MRE activity, highlighting the importance of local habitat continuity in creating effective corridors for bat movement. Woody plant diversity was also positively associated with SRE activity. These results were mostly consistent across simplified open croplands and structurally complex bocage landscapes.
Synthesis and Applications: Our findings underscore the value of targeted hedgerow management in promoting on-farm bat activity, and support the integration of simple hedgerow quality assessment protocols into agri-environment schemes. Consistent with previous research, we confirm the importance of tall, tree-filled hedgerows and emphasize the need to preserve mature trees. We also provide novel insights: hedgerow continuity and low-intensity trimming methods can further improve habitat suitability, particularly for tree-dwelling bats. Our management recommendations align with evidence from other taxonomic groups, including birds and invertebrates, and can contribute to broader biodiversity conservation strategies.
Dataset DOI: 10.5061/dryad.9s4mw6mvz
Description of the data and file structure
Files and variables
File: Dataset_hegderow_bat_activity_2025.csv
Description:
Using passive acoustic monitoring, this study aimed to disentangle the relative effects of six hedgerow characteristics (i.e. trimming method, height, width, foliage density, connectivity and woody plant diversity) on bat activity.
We used data from passive acoustic monitoring of bats along 448 hedgerows across three French regions differing in landscape configuration and composition. Hedgerows were selected to cover contrasting structural characteristics while ensuring consistent experimental conditions (adjacent to crop fields, away from artificial lighting and major infrastructures). Bat echolocation calls were recorded over 1,227 nights between 2020 and 2023 using passive acoustic devices (Song Meters and AudioMoths), following the French National Bat Monitoring Programme guidelines (e.g. under homogenous and favourable weather conditions). Acoustic data were processed with automated identification software (Tadarida) and grouped into three foraging guilds (short-, mid-, and long-range echolocators), with additional focus on two key species strongly associated with hedgerow landscapes (Rhinolophus ferrumequinum, Barbastella barbastellus). Concurrently, hedgerow structural and management characteristics were recorded using a simple standardised field protocol over a 50 m segment. All surveys were conducted with landowner and farmer permission.
As acoustic sampling cannot differentiate individual bats, we used bat activity—the number of bat passes per site per night—as a proxy for abundance. A bat pass was defined as a five-second sequence containing at least one echolocation call. As manual classification would have been prohibitively time consuming, we used automated species identification with the Tadarida software which assigned each bat call to the most accurate taxonomic level along with a confidence index. We excluded data activity data with a maximum confidence index below 0.5, reducing false positives while retaining a high number of passes. Then, we validated our statistical results using a more conservative threshold of 0.9, which further limited false positives but excluded more true positives. Thus, our dataset includes bat activity variables computed at both thresholds (0.5 and 0.9), denoted by the suffixes "05" and "09," respectively.
Variables:
- hedgerow_ID: Identifier code for each 50 m hedgerow segment.
- lat_y: Geographic coordinates (latitude) in decimal degrees (WGS84).
- long_x: Geographic coordinates (longitude) in decimal degrees (WGS84).
- study_region: Study region (3 levels: "East", "West", "North").
- observer: Initials of the observer who conducted the hedgerow survey and set up the passive acoustic monitoring device.
- date: Date of bat monitoring, i.e date at the start of the recording (recording spans from 30 after sunset to 30 after sunrise).
- julian_day: Julian day.
- device_type: Type of acoustic monitoring device used (2 levels: "AM" = Audiomoth, "SM" = SongMeter).
- vigiechiro_ID: Acoustic data identifier from the Vigie-Chiro citizen science monitoring programme archived by the French National History Museum (MNHN).
- recording_duration: Acoustic recording duration in hours (from 30 min after sunset to 30 after sunrise).
- night_length: Night duration in hours from sunset to sunrise on the date of bat monitoring.
- cluster_ID_3km: Cluster variable grouping all hedgerows located less than 3 km from one another.
- cluster_ID_5km: Cluster variable grouping all hedgerows located less than 5 km from one another.
- cluster_ID_8km: Cluster variable grouping all hedgerows located less than 8 km from one another.
- height: Hedgerow height in meters, categorical variable (3 levels: "< 5 m", "5-7 m", "> 7 m").
- width: Hedgerow width in meters, categorical variable (2 levels: "< 2 m", "> 2 m").
- woody_plant_diversity: Hedgerow genera-level woody plant diversity, categorical variable (2 levels: "low" = less or equal to two genera, "high" = more than two genera).
- trimming_method: Hedgerow trimming method, categorical variable (3 levels: "coppicing", "pollarding", "vertical_trimming").
- foliage_density: Hedgerow foliage density, categorical variable (2 levels: "low" = improved manoeuvrability for bat flight through the vegetation, "high" = reduced manoeuvrability for bat flight through the vegetation).
- connectivity: Hedgerow connectivity, categorical variable (3 levels: "0" = isolated hedgerow, "1" = hedgerow connected at one extremity, "2" = hedgerow connected at both extremities).
- dist_nearest_water_point_m: Hedgerow distance to nearest water point (e.g. pond, lake, river, stream) in meters.
- dist_nearest_forest_m: Hedgerow distance to nearest forest in meters.
- crop_cover_1000m: Crop cover (%) within a 1,000 m buffer radius.
- crop_cover_1500m: Crop cover (%) within a 1,500 m buffer radius.
- grassland_cover_1000m: Grassland cover (%) within a 1,000 m buffer radius.
- grassland_cover_1500m: Grassland cover (%) within a 1,500 m buffer radius.
- forest_cover_1000m: Forest cover (%) within a 1,000 m buffer radius.
- forest_cover_1500m: Forest cover (%) within a 1,500 m buffer radius.
- hedgerow_density_1000m: Hedgerow linear density (km/km²) within a 1,000 m buffer radius.
- hedgerow_density_1500m: Hedgerow linear density (km/km²) within a 1,500 m buffer radius.
- river_density_1000m: River linear density (km/km²) within a 1,000 m buffer radius.
- river_density_1500m: River linear density (km/km²) within a 1,500 m buffer radius.
- crop_cover_3000m: Crop cover (%) within a 3,000 m buffer radius.
- grassland_cover_3000m: Grassland cover (%) within a 3,000 m buffer radius.
- forest_cover_3000m: Forest cover (%) within a 3,000 m buffer radius.
- hedgerow_density_3000m: Hedgerow linear density (km/km²) within a 3,000 m buffer radius.
- river_density_3000m: River linear density (km/km²) within a 3,000 m buffer radius.
- LRE_activity_50: Sum of bat activity for all detected long-range echolocators (LRE; i.e. Eptesicus spp., Nyctalus spp.), based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- LRE_activity_90: Sum of bat activity for all detected long-range echolocators (LRE; i.e. Eptesicus spp., Nyctalus spp.), based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- SRE_activity_50: Sum of bat activity for all detected short-range echolocators (SRE; i.e. Barbastella spp., Plecotus spp., Rhinolophus spp., Myotis spp.), based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- SRE_activity_90: Sum of bat activity for all detected short-range echolocators (SRE; i.e. Barbastella spp., Plecotus spp., Rhinolophus spp., Myotis spp.), based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Barbar_09: Barbastella barbastellus activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Eptser_09: Eptesicus serotinus activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Myo_alc_cap_dau_ema_mys_09: Myotis spp. activity (i.e. Myotis alcathoe/capaccinii/daubentonii/emarginatus/mystacinus), based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- MyoGT_09: Myotis myotis activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Myonat_09: Myotis nattereri activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Nyc_lei_noc_09: Nyctalus leisleri/noctula activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Pip_nat_kuh_09: Pipistrellus kuhlii/nathusii activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Pippip_09: Pipistrellus pipistrellus activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Plec_aus_aur_09: Plecotus austriacus/auritus activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Rhifer_09: Rhinolophus ferrumequinum activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Rhihip_Rhieur_09: Rhinolophus hipposideros/euryale activity, based on automated species identification with a 0.9 (90%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Barbar_05: Barbastella barbastellus activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Eptser_05: Eptesicus serotinus activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Myo_alc_cap_dau_ema_mys_05: Myotis spp. activity (i.e. Myotis alcathoe/capaccinii/daubentonii/emarginatus/mystacinus), based on automated species identification with a 0.5 (50%)confidence threshold using the TADARIDA software (MNHN, 2017).
- MyoGT_05: Myotis myotis activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Myonat_05: Myotis nattereri activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Nyc_lei_noc_05: Nyctalus leisleri/noctula activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Pip_nat_kuh_05: Pipistrellus kuhlii/nathusii activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Pippip_05: Pipistrellus pipistrellus activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Plec_aus_aur_05: Plecotus austriacus/auritus activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Rhifer_05: Rhinolophus ferrumequinum activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- Rhihip_Rhieur_05: Rhinolophus hipposideros/euryale activity, based on automated species identification with a 0.5 (50%) confidence threshold using the TADARIDA software (MNHN, 2017).
- min_temperature: Lowest temperature (°C) during the night of the recording.
- cumulated_rainfall: Cumulative rainfall (mm) recorded on the day corresponding to the start of the night (day n) and the following day corresponding to the end of the night (day n + 1).
- mean_wind_speed: Averaged wind speed (km/h) during the night of the recording.
