Providing evidence to improve practice

Action: Use shelterwood cutting instead of clearcutting Bat Conservation

Key messages

  • One site comparison study in North America found higher or equal activity of at least five bat species in shelterwood harvests than unharvested control sites.
  • One replicated, site comparison study in Australia found Gould’s long eared bats selectively roosting in shelterwood harvests, but southern forest bats roosting more often in mature unlogged forest. A replicated site comparison study in Italy found barbastelle bats favoured unmanaged woodland for roosting and used shelterwood harvested woodland in proportion to availability.

Supporting evidence from individual studies


In a replicated, site comparison study in July and August 2001–2002 in beech Fagus sylvatica woodlands in central Italy (Russo et al 2004), barbastelle bats Barbastella barbastellus were found to use shelterwood harvested woodland for roosting in proportion to availability, whereas unmanaged woodland was positively selected, and pasture interspersed with woodland was avoided (four roosts in 69 ha of shelterwood, 19 roosts in 250 ha of unmanaged woodland and 10 roosts in 381 ha of pasture). Shelterwood harvested stands were selectively harvested (understory trees only) to reduce competition among trees in 1997–2002. Unmanaged woodland had not been logged for at least 40 years. Bats were caught using mist nets near cattle troughs used by bats for drinking, and fitted with radio transmitters. Twenty-five adult barbastelle bats were followed to 33 roosts where roost characteristics were recorded and compared to random trees. Most roost trees were dead, and were significantly taller, with a larger diameter and more cavities than random trees.


A site comparison study in May–September 2006 in two oak-hickory forests Carya spp. in southern Ohio, USA (Titchenell et al 2011) found activity of three bat species (red bat Lasiurus borealis, big brown bat Eptesicus fuscus, and silver-haired bat Lasionycteris noctivagans) in shelterwood harvested sites but not in unharvested control sites (50% shelterwood: red bat average 1.1 passes/site, big brown bat and silver-haired bat combined average 0.6 bat passes/site; 70% shelterwood: red bat average 0.6 passes/site, big brown bat and silver-haired bat combined average 0.8 bat passes/site; control sites: 0 bat passes). There was no difference in the activity of Myotis spp. and tri-colored bats Perimyotis subflavus between harvested sites (50% shelterwood: combined average 0.2 passes/night, 70% shelterwood: 0.24 passes/night) and unharvested control sites (combined average 0.07 passes/night). There was no difference in activity for any of the species between two harvesting retention levels (50% and 70% full stocking). Bat activity declined as structural volume in the understorey to the mid-canopy of the shelterwood harvests increased. The average number of bat passes decreased by 50% at volumes exceeding 17 m3/ha within 3–6 m of the forest floor. Estimated use by red bats decreased by 50% when volumes within 0–12 m exceeded 1750 m3/ha, and estimated use by big brown bats and silver-haired bats was highest when volumes within 3–6 m exceeded 63 m3/ha. In each study area two replicates of a control and two shelterwood harvests (50% and 70% stocking levels) of 10 ha were surveyed. In the shelterwood harvests a combined crown thinning and low thinning, favouring dominant and co-dominant oak, was done from March 2005 to June 2006 to reduce the stocking level. A random selection of two of the study sites were sampled per night with each site sampled for 6–8 nights. The vegetation was sampled in eight plots within each site. Bat activity was recorded for three hours from 30 minutes before sunset in the same eight plots using bat detectors. Mist netting was conducted during June–August 2006 for one to two nights per week.


In a replicated, site comparison study in February–March 2009 in jarrah Eucalyptus marginata forests in south-western Australia (Webala et al 2010), tracked Gould’s long-eared bats Nyctophilus gouldi were found to selectively roost in remnant trees in shelterwood harvests (10 bats, 37%). The remainder of tracked Gould’s long-eared bats roosted in gap release systems (one bat, 3%), mature forest (eight bats, 30%) and riparian buffers (eight bats, 30%). Only one southern forest bat Vespadelus regulus tracked during the study roosted in shelterwoods. Southern forest bats roosted more in mature unlogged forest (15 bats, 71%) and riparian buffers (five bats, 24%). Bats changed roosts every 1–2 days but showed fidelity to a general area. Shelterwoods had retention levels of 40–60%. Gap release systems involved the removal of 95% of the mature overstory. Riparian buffers and mature forest areas had been undisturbed for more than 30 years with only light selective logging prior to this. Bats (mostly females) were caught at two different water holes with different logging histories and fitted with radio transmitters. Ten southern forest bats were successfully tracked to 21 different roosts for an average of six days. Eleven Gould’s long-eared bats were successfully tracked to 27 roosts for an average of four days. Characteristics of roost trees were recorded and compared to random trees, and roost availability in the surrounding forest landscape was estimated.

Referenced papers

Please cite as:

Berthinussen, A., Richardson, O.C., Smith, R.K., Altringham, J.D. & Sutherland, W.J. (2017) Bat Conservation. Pages 67-93 in: W.J. Sutherland, L.V. Dicks, N. Ockendon & R.K. Smith (eds) What Works in Conservation 2017. Open Book Publishers, Cambridge, UK.