Retain remnant forest or woodland on agricultural land
Overall effectiveness category Likely to be beneficial
Number of studies: 6
Background information and definitions
There is evidence that remnant forest fragments provide important habitat for bats in agricultural landscapes (e.g. Struebig et al 2008, Lentini et al 2012). For studies that may carry out this intervention alongside other interventions to benefit bats on farmland, see ‘Introduce agri-environment schemes’. For a general intervention that involves retaining remnant habitats, see ‘Habitat protection – Retain remnant habitat patches’.
Lentini P.E., Gibbons P., Fischer J., Law B., Hanspach J. & Martin T.G. (2012) Bats in a farming landscape benefit from linear remnants and unimproved pastures. PLoS ONE, 7, e48201.
Struebig M.J., Kingston T., Zubaid A., Mohd-Adnan A. & Rossiter S.J. (2008) Conservation value of forest fragments to Palaeotropical bats. Biological Conservation, 141, 2112–2126.
Supporting evidence from individual studies
A replicated, site comparison study in 1999 of four agricultural sites with remnant forest and plantations in Western Australia (Hobbs et al 2003) found that remnants of native forest had higher overall bat activity and more bat species than plantations or agricultural grazing land. More bat passes were recorded in remnant forest (75 bat passes) than in plantations next to remnant vegetation (52 bat passes), isolated plantations (4 bat passes) or over agricultural grazing land (14 bat passes), although no statistical tests were carried out. More bat species were recorded in remnant forest (8 species) than in plantations and grazing land (2–4). Eight bat species were recorded in total (see original reference for data for individual species). All four sites had remnants of original native forest, farm forestry plantations (4–6 year old native bluegum Eucalyptus globulus), and open grazing land. At each of four sites, one location within each of four habitats (remnant forest, plantations next to remnants, isolated plantations and grazing land) was sampled with a bat detector for one full night in October 1999.Study and other actions tested
A replicated, site comparison study in 2002 of 120 sites in an agricultural area in New South Wales and Victoria, Australia (Law & Chidel 2006) found that remnants of original forest had higher bat activity than plantations or treeless grazed paddocks, but the number of bat species did not differ. More bat passes were recorded in remnants of original forest (average 302 bat passes/night) than in plantations (87 bat passes/night) or treeless grazed paddocks (50 bat passes/night). However, a similar number of bat species were recorded in remnants of forest (7 species), plantations (5–7 species) and paddocks (5 species). Eleven bat species were recorded in total (see original reference for data for individual species). Grazing land with small remnants of forest had been planted with native tree species from the mid-1970s to 1991. There were twelve treatments including different shapes or sizes (narrow, small, medium, large, very large) and ages (<10 or >10 years old) of remnants of original forest, plantations, and grazed paddocks with and without trees. For each of twelve treatments, 10 points were sampled with bat detectors for one full night in November–December 2002.Study and other actions tested
A replicated study in 2002 of 12 agricultural sites in southeastern Australia (Lumsden & Bennett 2005) found that remnants of native woodland had higher activity for three bat species than paddocks without trees, but no difference was found for seven other bat species or species groups. Average bat activity was significantly higher in remnants of woodland than treeless paddocks for the chocolate wattled bat Chalinolobus morio (remnants: 2.3 bat passes; treeless paddocks: 0.3 bat passes), western broad-nosed bat Scotorepens balstoni (remnants: 0.9 bat passes; treeless paddocks: 0.1 bat passes) and little forest bat Vespadelus vulturnus (remnants: 36 bat passes; treeless paddocks: 2 bat passes). There was no significant difference between remnant woodland and treeless paddocks for the white-striped free-tailed bat Tadarida australis, southern free-tailed bat Mormopterus spp., eastern free-tailed bat Mormopterus spp., Gould’s wattled bat Chalinolobus gouldii, the large forest bat Vespadelus darlingtonia, the southern forest bat Vespadelus regulus or long-eared bats Nyctophilus spp. (see the original reference for detailed results for each species). Two remnant woodlands (<10 ha) and two treeless paddocks were surveyed in each of three study areas. Each of 12 sites was sampled with a bat detector for four nights in January–April 2002.Study and other actions tested
A replicated, site comparison study in 2006–2007 at 14 farms in Tasmania (Law et al 2011) found that remnant native woodlands had higher bat activity, more bat species and more bat roosts than plantations and treeless paddocks, and species composition also differed. Bat activity and the number of bat species recorded was higher in remnant woodland (650 bat passes/night, 10 species) than in plantations (87 bat passes/night, 6–8 species) and paddocks (40 passes/night, 7 species), although no statistical tests were carried out. Twenty-eight bat roosts were identified in remnant trees, but none in plantations. Species composition differed in remnant woodland compared to plantations and paddocks (data reported as results of statistical models). Twelve bat species were recorded in total (see original reference for data for individual species). Forty-four sites were surveyed across 14 farms (11 in remnant woodland, 27 in plantations, six in treeless paddocks). Plantations (2–40 ha) consisted of 1–4 Eucalyptus spp. and were 4–5 or 10 years old. Each of 44 sites was surveyed for two consecutive nights/site in September 2006 and February 2007. Ten bats were caught in harp traps and radiotracked in late summer and spring 2008 at three farms.Study and other actions tested
A replicated, site comparison study in 2009–2010 of 34 woodland patches in an agricultural landscape in central Scotland, UK (Fuentes-Montemayor et al 2013) found that remnant woodland patches had higher activity for three bat species or species groups than surrounding pasture or arable land. The average number of bat passes recorded was significantly higher in the interior of remnant woodland patches than over pasture or arable land for common pipistrelle Pipistrellus pipistrellus (woodland: 0.9; pasture/arable: 0.1), soprano pipistrelle Pipistrellus pygmaeus (woodland: 22; pasture/arable: 2) and Myotis spp. (woodland: 0.1; pasture/arable: 0.02). All of 34 broadleaved or mixed woodland patches (0.1–30 ha) were >60 years old and surrounded by pasture and/or arable land. At each of 34 woodland patches, bat activity was recorded with bat detectors for 10 minutes at 2–4 sampling points in the woodland interior and two sampling points over pasture/arable land >20 m from the woodland edge. Each of 34 sites was surveyed once in June–August 2009 or May–July 2010.Study and other actions tested
A replicated, site comparison study in 2010–2011 in eight sites of remnant forest and plantations in an agricultural area of central Brazil (Pina et al 2013) found that remnant forest fragments had more bats and bat species captured within them than plantations, but species composition did not differ. Significantly more bats and more bat species were captured in remnant forest fragments (134 bats of 13 species) than in plantations (75 bats of 8 species). However, species composition did not differ significantly (data reported as statistical model results). Four fragments (150–378 ha) of each of two habitat types were surveyed: Eucalyptus spp. plantations and semi-deciduous native remnant forest. Fragments were surrounded by soybean or sugar cane plantations and cattle pastures. Each of eight fragments was sampled for one night/month between December 2010 and April 2011 with 10 mist nets deployed along linear transects for four hours from sunset.Study and other actions tested