Clear or open patches in forests
Overall effectiveness category Evidence not assessed
Number of studies: 14
Background information and definitions
Although many species of butterflies and moths live within woodland, many are reliant upon open glades, where sunlight can reach the understorey and create warm habitat patches and encourage the growth of flowers. Traditional woodland management, such as small scale wood harvesting, which created these habitats has declined, leading to the closure of woodland canopies (Bubová et al. 2015). Opening up patches within the forest may create new habitat for these species. However, some species, especially in tropical and sub-tropical regions, are dependent on the shade provided by continuous forest cover, and may be adversely affected by canopy clearance, so caution should be taken when applying this action.
Bubová T., Vrabec V., Kulma M. & Nowicki P. (2015) Land management impacts on European butterflies of conservation concern: a review. Journal of Insect Conservation, 19, 805–821.
Supporting evidence from individual studies
A replicated, site comparison study in 1980–1984 in 12 woodlands in Kent, UK (Warren 1987) found that woodland patches cleared two to four years ago had a higher abundance of heath fritillary Mellicta athalia than areas cleared more recently or longer ago. The number of heath fritillary recorded 2–4 years after a site had been cleared (108–410 individuals/site) was higher than in the first year after clearance (9 individuals/site), or 7–11 years after clearance (2–17 individuals/site). Some populations became extinct nine years after management. Twelve woodlands were managed by clearing patches to plant conifers, 1–11 years before surveying. From 1980–1984, the number of adult heath fritillary at each site was estimated from a combination of counts on regular transects throughout the season, single counts around the peak flight period, and mark-recapture of individuals.Study and other actions tested
A replicated, site comparison study in 1986 in a mixed woodland in Dorset, UK (Robertson et al. 1988) found that managed clearings within a woodland had a higher abundance and species richness of butterflies than other areas of the wood. In managed clearings, both the abundance (89 individuals/km) and species richness (19 species) of butterflies were higher than in coppiced woodland (abundance: 25 individuals/km; richness: 16 species), unmanaged broadleaved woodland (abundance: 2 individuals/km; richness: 4 species), or conifer plantations (abundance: 5 individuals/km; richness: 2 species). See paper for individual species results. The woodland contained patches managed in four ways: managed clearings (30–50 m wide and 100–150 m long) which were cleared of scrub every three years; open woodland with coppiced hazel; unmanaged broadleaved woodland with unmanaged hazel coppice; and conifer plantation. In July–August 1986, butterflies were surveyed six times on each of twenty-two 200-m transects: four in managed clearings, eight in open, coppiced wood, six in unmanaged wood and four in conifer plantation.Study and other actions tested
A study in 1992–1996 in a forest in Ruokolahti, Finland (Marttila et al. 1997) reported that, in an area with selected clearance of Scots pine Pinus sylvestris, a translocated population of baton blue Pseudophilotes baton schiffermuelleri butterflies increased in size and survived for at least two years. One year after a population of 10 baton blue females were released, 24 butterflies were captured, and two years after release, 46 butterflies were captured. The 230 x 70-m Scots pine-covered site in Ruokolahti was divided into 10-m2 study plots. In 1992, pines were selectively logged, and in 1996 some additional saplings were removed. In Säkylänharju, in June 1994, ten female baton blues were captured and marked before being translocated 315 km east to Ruokolahti where they were released into different 10-m2 plots. In June 1995 and June and July 1996, surveys were conducted twice/day along the longitudinal boundaries of all plots, during which butterflies were captured, marked and released, and egg-laying and mating sightings were recorded.Study and other actions tested
A before-and-after study in 1987–1999 on a rocky island in South Karelia, Finland (Marttila et al. 2000) reported that following selected clearance of pine trees, chequered blue butterfly Scolitantides orion numbers increased. No statistical tests were carried out. In the three years prior to clearance around two chequered blue butterflies were recorded annually, later in the year of clearance none were recorded, and in the nine years after clearance 0–25 were recorded annually. In 1990, pines were selectively logged on the rocky outcrops and slopes of an island on Lake Saimaa, creating a patchy, open habitat. In 1990–1996 and 1998–1999 chequered blue abundance was estimated on long (2–5 hours) and short (≤ 30 minutes) surveys 1–3 times annually on days with good weather during the peak flight period. In June 1997, butterflies were monitored by an eight-day mark-recapture study. Authors reported that butterflies were also observed annually in 1987–1989 but provided no survey details.Study and other actions tested
A replicated, controlled, before-and-after study in 1992–1997 in 18 deciduous woods in Östergötland, Sweden (Bergman 2001) found that clearing new woodland glades increased the population size of woodland brown Lopinga achine butterflies in five out of six woods. Over 4–5 years after glades were created, the population of woodland brown increased by 93–97% in five woods where glades were created, but decreased by 19–25% in nine woods where no glades were created. However, in a sixth wood cleared later, the population decreased by 27% in the first two years after glades were created, compared to a 9% decrease over the same time period in three other woods where no glades were created. The authors noted that populations only occurred at sites with >60% canopy cover, but sites with 70–75% canopy cover had the highest population density. From 1992–1995, irregularly-shaped glades (10–30 m long) were created in six woodlands (20 in one wood, and 5–6 in each of the others). Where possible, the longest side had a south-west to north-east orientation to maximize sun exposure. In July 1992–1997, the adult population size of woodland brown butterflies was estimated in six woods where glades were created, and in 12 woods where no glades had been created.Study and other actions tested
A paired sites study in 1999 in a pine forest in Arizona, USA (Meyer and Sisk 2001) found that orange sulphur Colias eurytheme butterflies, but not pine white Neophasia menapia butterflies, flew into areas with selective clearance of ponderosa pine Pinus ponderosa more often than into areas without selective clearance. From both east- and west-facing edges, orange sulphur butterflies flew into an area of forest restored through selective logging more often than they flew into unrestored forest (58–90%). Pine whites did not fly into the restored area more often than expected by chance from either edge (47–57%). In early 1999, eighty-seven percent of trees in the restoration area were cut and removed. In unrestored areas no trees were cut. In July and August 1999, orange sulphur and pine white butterflies were collected from the study sites and nearby land, and held overnight at 5 °C. Thirty minutes before dawn, butterflies were released at four points 1 m from the east- or west-facing edge between the restored and unrestored areas (two points on the east- and two on the west-facing edge of the restored area). Orange sulphurs were monitored on three mornings and pine whites on one morning. Behaviour was monitored until they moved >10 m from the release point.Study and other actions tested
A replicated, controlled study in 2001–2004 in an urban evergreen forest in the Kansai region, Japan (Yamamoto & Tuhara 2004) found that small cleared patches in the forest had a higher abundance and species richness of butterflies than the forest interior, but both abundance and species richness decreased with time since clearing. Two to three years after clearing, the abundance (56–142 individuals) and species richness (14–19 species) of butterflies in cleared patches was higher than in uncleared patches in the forest interior (abundance: 11 individuals; richness: 9 species). However, in the four patches cleared in the first year, abundance and species richness were higher in the year after clearing (abundance: 161 individuals; richness: 20 species) than three years after clearing (abundance: 76 individuals; richness: 18 species). In 2001 and 2002, four patches/year (15 × 15 m each) were cleared within a mature ring-cupped oak Quercus glauca and Japanese bay tree Machilus thunbergii forest (0.16–0.20 trees/m2). From April–September 2004, butterflies were surveyed three times/month in each cleared patch, and in four nearby patches of forest interior, for 10 minutes/plot. The four patches cleared in 2001 were surveyed in the same way in 2001.Study and other actions tested
A replicated, before-and-after, site comparison study in 2006–2011 in woodlands in Hampshire, Wiltshire and Kent, UK (Hoare et al 2012) reported that in areas where patches of forest were cleared, along with cutting, grazing and ride-widening, the number of pearl-bordered fritillary Boloria euphrosyne and Duke of Burgundy Hamearis lucina breeding sites increased, whereas in a nearby area with no management pearl-bordered fritillary breeding sites decreased. In 2006, there were five pearl-bordered fritillary breeding sites in Tytherley Woods, and in 2011, after habitat management, there were 12 sites, and most populations were stable or increasing (details not provided). However, in nearby woodland without management there were 10 pearl-bordered fritillary breeding sites in 2006 and seven in 2011. In Denge Woods in 2007 there were 11 Duke of Burgundy butterflies recorded across two sites, and in 2010 after habitat management there were 173 butterflies across 10 sites (although two of these sites had not been surveyed in 2007). Results were not tested for statistical significance. In 2007–2011 habitat management was conducted in Tytherley Woods, Hampshire/Wiltshire and Denge Woods, Kent, which included clearing derelict coppice, felling non-native conifer plantations, cutting and grazing clearings and widening rides. Butterfly surveys were conducted in 2006–2010 in sites which pearl-bordered fritillary and Duke of Burgundy butterflies were known to occupy and in other suitable habitat patches.Study and other actions tested
Referenced paperHoare D., Dent K., Kelly C., McLellan L., Thompson F. & Wheatley S. (2012) Landscape-scale woodland restoration for multiple species in the South East Woodlands. Pages 66-75 in: S. Ellis, N.A. Bourn & C.R. Bulman (eds.) Landscape-scale conservation for butterflies and moths: lessons from the UK. Butterfly Conservation.
A before-and-after study in 2002–2011 in an area of woodland, heathland and grassland in South Wales, UK (Hobson & Smith 2012) reported that after tree felling and scrub clearance was conducted, along with coppicing and grazing, the numbers of high brown fritillary Argynnis adippe and small pearl-bordered fritillary Boloria selene butterflies increased. Results were not tested for statistical significance. In 2002, prior to felling and scrub control, an average of seven high brown fritillary adults/hour were recorded at the site, and seven years after felling and scrub clearance began 14 adults/hour were recorded. Small pearl-bordered fritillary also increased in number between 2003 and 2011 (numbers not given). From 2003–2011 some large trees were felled and some patches of bracken Pteridium aquilinum, bramble Rubus fruticosus agg., honeysuckle Lonicera periclymenum and tree saplings were cut back in the Alun Valley, a 254 ha landscape comprising woodland, heathland and grassland. Coppicing of hazel Corylus avellana and gorse Ulex europaeus (1999–2011) and sheep grazing (start date not given–2011) were also conducted in areas. Adult butterflies were counted annually from 2002–2011.Study and other actions tested
A replicated, site comparison study in 2010 in six deciduous woodlands in Hampshire and Wiltshire, UK (Merckx et al. 2012) found that wide woodland rides had a lower abundance, but similar species richness, of macro-moths than standard rides and mature forest. In wide woodland rides, the abundance of macro-moths (1,926 individuals) was lower than in standard rides (2,513 individuals) and mature forest (2,479 individuals). Species richness was similar between wide (175 species) and standard (176 species) rides and mature forest (180 species). However, wide rides and coppiced woodland supported 49 species not found in standard rides or mature forest, and 124 species were more abundant in wide rides and coppiced woodland than in standard rides and mature forest, especially ‘common but severely declining’ species (see paper for details). Only 22 species were found in standard rides or mature forest but not wide rides or coppiced woodland. Within six woodlands (8–711 ha), six areas under each of six management types were studied: young (1–2 years), medium (3–6 years) and old (7–9 years) coppice, wide (>20 m) and standard (<10 m) rides, and non-coppiced mature forest. From July–October 2010, macro-moths were sampled nine times/site using a 6 W Heath actinic light trap, over 27 nights (two sites/management type sampled/night).Study and other actions tested
A review in 2015 of 126 studies in Europe (Bubová et al. 2015) reported that maintaining sparse forest stands by clearing small patches benefitted 19 out of 67 butterfly species of conservation concern. Results were not tested for statistical significance. The review reported that 23 studies found that clearing and maintaining open areas in woodland benefitted 19 butterfly species. See paper for information on individual species. Clearings were created and maintained by felling trees, suppressing the growth of seedlings in glades, grazing forests, and coppicing. The review focussed on 67 butterfly species of conservation concern. The available information was biased towards studies in Northern and Western Europe.Study and other actions tested
A replicated, paired sites, controlled study in 2011–2012 in six oak/hornbeam forests in South Moravia, Czech Republic (Sebek et al. 2015) found that partially-cleared plots had higher butterfly but lower moth species richness than plots of closed-canopy forest, and species richness of butterflies and moths was affected by distance of the plot to the woodland edge. Butterfly species richness was higher in partially-cleared forest plots connected to the forest edge (16) and naturally open forest plots (14) than partially-cleared plots >20 m from the edge (10), which in turn had higher richness than closed-canopy forest plots (2). However, whilst moth species richness was higher in partially-cleared forest plots connected to the edge (97) and naturally open forest plots (111) than partially-cleared plots >20 m from the edge (81), all three had lower richness than plots of closed-canopy forest (130). In February 2011 and 2012, four 40-m2 plots were established at six forest sites: a partial clearing at a forest edge connected to a meadow, a partial clearing >20 m from forest edges, closed-canopy forest, naturally open forest. A few trees were left in the two clearings to replicate open forest. In the May–September after plot establishment, adult butterflies were surveyed five times for seven minutes/plot, and moths were captured once/month using one light trap/plot.Study and other actions tested
Referenced paperSebek P., Bace R., Bartos M., Benes J., Chlumska Z., Dolezal J., Dvorsky M., Kovar J., Machac O. & Mikatova B. (2015) Does a minimal intervention approach threaten the biodiversity of protected areas? A multi-taxa short-term response to intervention in temperate oak-dominated forests. Forest Ecology and Management, 358, 80-89.
A replicated, site comparison study in 2016 in 10 grasslands and forest clearings in Honshū, Japan (Ohwaki et al. 2018) found that clearcut forest patches attracted a similar number of butterfly species as semi-natural grasslands. In clearcut patches, the total number of butterfly species (22 species/site) and the number of threatened butterfly species (2 species/site) were not significantly different to semi-natural grasslands (total: 26 species/site; threatened: 6 species/site). However, 15 butterfly species (including five threatened species) were observed only in grassland, compared to six species (none threatened) which occurred only in clearcuts (statistical significance not assessed). From 2008–2012, five plantation patches (aged 27–88 years, 3.1–14.7 ha, >1.8 km apart) were clearcut. Two years after cutting, larch Larix kaempferi seedlings were planted at each site, with some broadleaved deciduous and evergreen coniferous trees. For 3–5 years after planting, summer mowing was used to suppress surrounding vegetation. Five semi-natural grasslands (15–1,900 ha, >3.5 km apart) were managed by burning in early spring (three sites) or had been abandoned since the 1950s (two sites). From May–October 2016, butterflies were surveyed once/month on three 200-m transects/site (six transects on the largest grassland).Study and other actions tested
A replicated, site comparison study in 2015 in a boreal forest in Alberta, Canada (Riva et al. 2018) found that larger cleared patches in forests had a higher abundance and species richness of butterflies than undisturbed forest, but smaller cleared patches did not. The abundance and species richness of butterflies in large clearings (abundance: 65 individuals/site; richness: 13 species/site) and wide corridors (abundance: 95 individuals/site; richness: 15 species/site) was higher than in undisturbed forest (abundance: 21 individuals/site; richness: 7 species/site). However, narrow corridors (abundance: 31 individuals/site; richness: 8 species/site) were similar to undisturbed forest. Of 43 species observed, 41 had a higher abundance in cleared sites than in undisturbed forest (statistical significance of individual species results not presented, see paper for details). From 2000–2005, clearings (60 × 60 m) and corridors (3 or 9 m wide) were created in a 25-km2 area of previously undisturbed forest by removing trees. From June–August 2015, butterflies were surveyed 11 times on five 200-m transects in each type of clearing, and in undisturbed forest patches which had received no wildfire or anthropogenic disturbance within 50 m for >80 years.Study and other actions tested