Install underpasses or culverts as road/railway crossing structures for bats
-
Overall effectiveness category Likely to be beneficial
-
Number of studies: 8
View assessment score
Hide assessment score
How is the evidence assessed?
-
Effectiveness
-
Certainty
-
Harms
Study locations
Supporting evidence from individual studies
A study in 2004–2007 of an underpass below a motorway in a forested area of northern Bavaria, Germany (Kerth & Melber 2009) found that a cluttered habitat bat species rarely crossed the motorway and only crossed through the underpass, whereas an open habitat bat species crossed the motorway frequently and flew over the road more often than through the underpass. Only three of 34 radio-tracked Bechstein’s bats Myotis bechsteinii crossed the motorway, all using the underpass (36 crossings). Five of six radio-tracked barbastelle bats Barbastella barbastellus crossed the motorway but flew over the road (21 crossings at six different sites) more often than through the underpass (16 crossings). The motorway had four to five lanes carrying an average of 84,000 vehicles/day. The underpass (5 m wide x 4.5 m high x 30 m long) was located within a motorway section surrounded by forest. Mist netting was carried out for 153 nights at 12 sites within the forest in May–September 2004–2007. Each of 40 adult female bats was radio-tracked for at least three full consecutive nights.
Study and other actions testedA replicated, site comparison study in 2008 at 13 under-motorway crossing routes in agricultural and woodland habitat in southern Ireland (Abbott et al 2012) recorded more bat activity in under-motorway routes (underpasses or rivers bridged over by the road) than over the road above them, or in adjacent habitats. More bats were recorded in under-motorway routes (underpasses: 662 bat passes; river bridges: 4,692 bat passes) than over the road above them (above underpasses: 45 bat passes; above river bridges: 96 bat passes). Bat activity was also greater (by >10%) at under-motorway crossing routes than in adjacent habitats (data reported as statistical measures). Six bat species or species groups were recorded in total (see original paper for data for individual species). The motorway (65–70 m wide) had four lanes carrying an average of 20,000 vehicles/day. Seven underpasses (5–17 m wide x 4–10 m high x 26–63 m long) and six river bridges (6–420 m wide x 3–19 m high x 23–39 m long) were surveyed. Bat detectors recorded bat activity above and below each of the 13 structures and simultaneously at two adjacent linear features on two nights in May–September 2008.
Study and other actions testedA study in 2009 at an underpass below a motorway in an agricultural area of Ennis, west Ireland (Abbott et al 2012) found that a large underpass was used by five of six bat species although 2–18% of bat passes were recorded over the road above the underpass. Two edge habitat adapted bat species (soprano pipistrelle Pipistrellus pygmaeus, common pipistrelle Pipistrellus pipistrellus) were recorded most frequently in the underpass (soprano pipistrelle: 770 bat passes; common pipistrelle: total 469 bat passes) but 18% of bat passes were recorded over the road above (soprano pipistrelle: 174 bat passes; common pipistrelle: 103 bat passes). The underpass was also used by cluttered habitat adapted brown long-eared bats Plecotus auritus (60 bat passes), lesser horseshoe bats Rhinolophus hipposideros (58 bat passes) and Myotis spp. (30 bat passes), with only a small number of bats recorded over the road above (1–3 bat passes, 2–5%). One open habitat adapted bat species, Leisler’s bat Nyctalus leisleri, was only recorded flying over the road above the underpass (56 bat passes). The motorway had four lanes carrying an average of 11,000 vehicles/day. The underpass (17 m wide x 6 m high x 26 m long) had a minor road through it. Simultaneous recordings were made with bat detectors above and below the underpass for 16 full nights in May 2009.
Study and other actions testedA replicated study in 2010 at three underpasses below two roads in an agricultural area of Cumbria, UK (Berthinussen & Altringham 2012) found that one of three underpasses had a greater proportion of bats flying through it than crossing over the road above at traffic height. At one underpass (6 m wide x 5 m high x 30 m long) located on an original bat commuting route, 96% of bats (864 of 904) flew through it to cross the road and 4% (32 of 904) flew over the road above at traffic height. At two underpasses (5 m wide x 2.5 m high x 15 m long; 6 m wide x 3 m high x 30 m long), 4% (39 of 1,117) and 31% of bats (11 of 36) flew through them and 67% (751 of 1,117) and 61% (22 of 36) crossed the road above at traffic height. Both underpasses were not located on original bat commuting routes, but attempts had been made to divert bats towards them with planting. The two roads had 2–3 lanes of traffic carrying an average of 12,000–17,000 vehicles/day. At each of three underpasses, crossing bats were observed and recorded with bat detectors during 10 x 90-minute surveys at dusk or dawn in June–July 2010.
Study and other actions testedA replicated study in 2013 at three underpasses below two roads in the UK (Berthinussen & Altringham 2015) found that more bats crossed through the underpasses than over the road above, but at two underpasses up to a third of bats still crossed the road above at traffic height. At one underpass (4.5 m wide and high x 45 m long), 95% of bats (608 of 639) flew through it to cross the road, and 5% (31 of 639) flew over the road above at traffic height. At two underpasses (2.5 m wide and high x 70 m long; 2.5 m wide and high x 45 m long), 70% (199 of 283) and 66% of bats (129 of 196) flew through them to cross the road, but 29% of bats (82 of 283 and 57 of 196) crossed the road above them at traffic height. Seven bat species or species groups were recorded in total (see original report for data for individual species). All three underpasses were installed for bats. Observations of crossing bats and recordings of bat calls were made during 6–10 x 60-minute surveys at dusk or dawn at each underpass in June–August 2013.
Study and other actions testedA replicated, site comparison study in 2013–2015 of six culverts and six open-span underpasses under a road in Victoria, Australia (Bhardwaj et al 2017) found that culverts and underpasses were used more frequently to cross the road by bat species adapted to cluttered habitats, but results were mixed for bat species adapted to open and edge habitats. Bat species adapted to cluttered habitats crossed the road more often through culverts (average 5 times/night) and underpasses (10 times/night) than over the road above (2 times/night above both). Bat species adapted to edge habitats crossed less often through culverts (1 time/night) than over the road above (13 times/night), but more often through underpasses (29 times/night) than over the road above (4 times/night). Bat species adapted to open habitats crossed more often over the road above culverts (31 times/night) and underpasses (19 times/night) than through culverts (1 time/night) or underpasses (12 times/night). Culverts were box culverts (3–3.6 m wide and high x 24–67 m long) with a concrete floor. Underpasses were large, open structures (10–90 m wide x 3–15 m high x 30–54 m long) with natural vegetation below. The road was a four-lane divided highway carrying an average of 8,000–14,000 vehicles/day. Six bat detectors were deployed/site to record crossing bats for a total of four full nights in December–January in 2013/2014 and 2014/2015.
Study and other actions testedA replicated, site comparison study in 2011–2013 and 2015 of eight structures under a road in Wales, UK (Davies 2019) found that larger underpasses and a ‘box’ culvert had higher activity of greater horseshoe bats Rhinolophus ferrumequinum within them than over the road above, but smaller ‘pipe’ culverts did not. Higher greater horseshoe bat activity was recorded within three large underpasses (68–97% of bat passes) and a ‘box’ culvert (88%) than over the road above (underpasses: 3–32%; box culvert: 12%), although the results were not tested for statistical significance. Three smaller ‘pipe’ culverts had lower greater horseshoe bat activity within them (15–40% of bat passes) than over the road above (60–85%). Three ‘pipe’ culverts (0.75, 1 and 1.5 m diameter), a ‘box’ culvert (1.8 x 3 m) and three underpasses (one for cattle: 2.7 x 2.4 m; two for horses: 3.5 x 4 m; one open-span: 8.95 x 3.5 m) were installed under a new road in 2011. All were on/near bat commuting routes or foraging habitat and 4.3–6.5 km from a greater horseshoe bat roost. In May–September, bat detectors within and above each structure recorded bat activity for one night/month (2011, 2012 and 2015) or on three occasions (2013). Surveyors using night-vision equipment confirmed bats were flying through or over the structures.
Study and other actions testedA replicated, site comparison study in 2018 of 24 underpasses along four roads in the Occitanie region, France (Laforge et al 2019) found that underpasses were used by five bat species or species groups in varying proportions, and use was influenced by underpass type and height, road width, and the amount of forest and hedgerows in the surrounding landscape. Myotis, Rhinolophus and Pipistrellus/Miniopterus spp. had greater activity in underpasses that bridged rivers than in culverts with rivers or roads through them (data reported as statistical model results). Taller underpasses were used less (compared to flying over road sections above or adjacent to them) than shorter ones by Rhinolophus spp., and Plecotus spp. were more likely to use underpasses under roads with more lanes (see original paper for details). Greater proportions of bats also used underpasses in landscapes with greater forest cover (Myotis spp. and barbastelle bats Barbastella barbastellus), fewer hedgerows (Myotis spp.) or that were closer to forest patches (Rhinolophus spp.). The 24 underpasses varied in width (2.5–15 m), height (2.5–6 m), length (5–100 m) and type (five river bridges, seven river culverts, 12 road or track culverts). The four roads were surrounded by a mixed landscape of woodland, hedgerows, grassland, and agricultural land. Bat detectors recorded bat activity within, above, and 200 m along the road from, each underpass for three consecutive full nights in July–August 2018.
Study and other actions tested
Where has this evidence come from?
List of journals searched by synopsis
All the journals searched for all synopses

This Action forms part of the Action Synopsis:
Bat Conservation
Bat Conservation - Published 2021
Update 2020