Noctule
A strongly tree-associated species that can use holes for breeding and winter shelter. Retaining several suitable trees matters because a colony may change roosts.
ACP bat ecology guide · Trees, roosts & better design
Understand the feature. Protect the roost. Design around the tree.
Bats in trees can be hidden behind a small bark plate or deep inside a branch cavity. This illustrated guide explains what those features mean, how an ecologist assesses them, and how a project can avoid harm before tree work begins.
01 · Start with the essentials
Yes. Many British bats use tree cavities, cracks and sheltered spaces as roosts. They do not excavate a nest or chew a new hole. Instead, they occupy features formed by decay, damage, woodpeckers or the tree’s growth.
A tree can provide shelter, insect prey and a connection to feeding areas. Different trees may offer different conditions as weather, breeding status and season change. A single occupied tree may therefore be part of a wider roosting resource.
This guide is for landowners, architects, arborists and project teams. It explains the ecological decisions behind tree work; it does not provide instructions for inspecting cavities or carrying out work around bats.
Read the evidence: BCT: roosts in trees.
02 · Illustrated feature atlas
An entrance may lead to a chamber that cannot be judged from its opening alone.
Damage and decay can leave sheltered space within a branch or trunk.
A narrow opening may continue further into the wood than is visible from below.
The sheltered space behind an attached bark plate can provide a crevice roost.
A hollow behind woundwood may offer space beyond the visible entrance.
A split at a fork can form a sheltered feature; not every fork contains a cavity.
Dense ivy can also obscure the underlying tree and complicate an assessment. Do not pull it away, lift bark or probe a hole to check. Record what can be observed safely and let a suitably experienced ecologist resolve the uncertainty.
Read the evidence: BCT: roosts in trees · BCT survey guidelines, fourth edition (2023; amended March 2024).
03 · Beyond the visible opening
The useful space is inside the feature. An entrance may open into an upward chamber, a long fissure or a shallow gap behind bark. Conversely, an impressive-looking opening may be wet, exposed or too shallow to offer suitable shelter.
Depth, shelter, wood thickness and exposure influence conditions within a feature. The surrounding canopy also changes sunlight and wind exposure. As a result, pruning neighbouring branches can alter a roost even when the entrance itself remains untouched.
A technical assessment records what was seen, what remains hidden and how confidently the feature can be described. Photographs help explain its location, but a ground photograph alone rarely resolves its internal extent.
Read the evidence: BCT survey guidelines, fourth edition (2023; amended March 2024) · BCT: roosts in trees.
04 · Species and ecological context
Species differ in their roost choices, and local availability matters. These examples explain ecological tendencies; they are not a field key for identifying the occupants of a particular tree.
A strongly tree-associated species that can use holes for breeding and winter shelter. Retaining several suitable trees matters because a colony may change roosts.
A woodland specialist associated with tree cavities, including woodpecker holes. A group of suitable trees and the woodland around them can be more important than one isolated feature.
Frequently associated with sheltered crevices and loose bark. Features can be short-lived, so maintaining a continuing supply of suitable trees is important.
Both can use trees as well as other roost types. Their woodland use also highlights the importance of feeding habitat and connected vegetation.
Pipistrelles and other species may use trees too. For species profiles, seasonal behaviour, anatomy and attributed recordings, explore our British bats guide.
Read the evidence: BCT: woodland specialists · Rachwald and colleagues (2025): tree and stand selection.
05 · What primary research adds
Russo and colleagues studied barbastelles in an old beech forest in central Italy. They found that females could retain social associations while moving between roosts. Group size and roost mates helped explain thermoregulation.
Practical implication: an apparently empty feature on one visit may still belong to a used roost network. Retention should consider alternative shelters and the routes between them.
Limit: this was a study of one species in a particular forest, not a universal timetable for British bats.
Read the evidence: Russo and colleagues (2017): sociality and roost switching.
Rachwald and colleagues radio-tracked 24 barbastelles and 13 soprano pipistrelles over three breeding seasons in Poland’s Białowieża Forest. They located 48 and 15 roosts respectively. Barbastelles mainly used dead spruce; pipistrelles predominantly used broadleaved trees.
Practical implication: examine the available features and local ecology rather than assuming that only a favoured tree species matters.
Limit: these findings describe a particular lowland forest and sampled animals. They do not provide a UK minimum tree diameter or a development buffer.
Read the evidence: Rachwald and colleagues (2025): tree and stand selection.
06 · Read the report correctly
The BCT fourth-edition framework separates the assessment of a tree from the suitability of a particular feature. That distinction makes reports clearer and prevents a potential maternity feature being mistaken for a confirmed maternity roost.
| Tree-level term | What it means | What the project team should understand |
|---|---|---|
| NONE | No potential features identified, or their presence judged very unlikely. | A recorded assessment with stated limitations, not a permanent guarantee. |
| FAR | Further assessment is required to establish whether features are present. | Uncertainty still needs resolving; it does not mean the tree is clear for work. |
| PRF | At least one potential roost feature is present. | The feature and the proposal need closer consideration. |
This feature category concerns suitability for an individual or very small number of bats, considering the feature and habitat context. It does not mean that bats are absent or that work is automatically permitted.
This category indicates suitability for multiple bats, potentially including a maternity group. It describes potential. Occupancy, species and roost function require supporting evidence.
An ecologist may revise an initial category after a closer inspection. The report should explain the evidence behind that change. Do not simply transfer the low, moderate and high building categories onto the tree framework.
Read the evidence: BCT survey guidelines, fourth edition (2023; amended March 2024).
07 · Methods should follow the question
The ecologist relates the trees to the proposal, examines visible features and records constraints. Leaf cover, height, ivy and viewing angles can limit what can be concluded. A map should connect each tree identifier to its features.
A closer inspection may clarify a feature’s extent and suitability. Access could involve trained climbing specialists or suitable access equipment, subject to safety, competence and any necessary licence. A fragile feature must not be damaged by the investigation.
Where uncertainty still matters to the proposal, the ecologist selects a proportionate method. This may involve repeated feature inspections, emergence observation or specialist landscape-scale work. Each method should answer a defined question.
Trees are not simply small buildings. The number of openings, changing foliage, canopy clutter and hidden exits can make observation difficult. An emergence survey may be appropriate where a feature cannot be inspected safely or fully, or to characterise a known roost.
An activity detector records bats passing its position. It does not establish that every nearby cavity is empty. Likewise, hearing bats beside a tree does not by itself prove that they emerged from that tree.
The report should distinguish an observed emergence, an inspected feature and an acoustic record nearby. If the method leaves a relevant area unseen, the limitation should remain visible in the conclusion. More visits are useful only if they improve the evidence needed for the decision.
Read the evidence: BCT survey guidelines, fourth edition (2023; amended March 2024).
08 · Proportionate evidence, explained
No. First establish whether impacts can be avoided. Where an impact remains, the approach depends on the feature, local context, scale of work and evidence already available. A blanket survey package is not a substitute for that assessment.
BCT describes a roost-resource approach for appropriate situations involving only PRF-I features. This combines advance compensation and a precautionary working method rather than automatically requiring further surveys. It is a professional decision, not a self-service clearance route; the number and landscape importance of the features still matter.
For PRF-M features, the guideline’s summer close-inspection approach is three visits between May and September, including at least two between May and August. Visits should normally be at least three weeks apart. Weather, geography, access and the ecological question can affect the design.
Read the evidence: BCT survey guidelines, fourth edition (2023; amended March 2024).
09 · Design-led advice
Often, this is the most useful question to ask early. A proposal that retains a tree, protects its features and preserves its connections may need a different evidence package from one that removes it. Avoidance must address indirect effects too.
These options need joint review by the designer, arborist and ecologist. Root-protection distances and bat-related measures answer different questions; one should not be treated as a substitute for the other.
See our design-options guide for the wider principle: understand the impact before committing to additional work.
10 · From one feature to the landscape
Keep the entrance and internal shelter functioning. Changes to the supporting branch, bark, canopy or nearby exposure can matter even where the trunk remains.
Review the vegetation and sheltered routes between the tree and feeding areas. A retained tree surrounded by new lighting or a widened open gap may function differently.
Consider future tree management, replacement planting and the development of new features. Young planting cannot immediately reproduce the shelter offered by old cavities.
A lighting brief should identify ecological receptors, operating periods and the surfaces or routes requiring protection. There is no single illuminance number that clears every tree-roost situation. Our bats and lighting guide explains how to coordinate the ecological brief and lighting design.
Read the evidence: Natural England: bats and planning decisions · BCT: woodland specialists.
11 · Seasons and survey limitations
| Period | What may be happening | Why the programme needs care |
|---|---|---|
| Spring | Bats become more active and may change roosts as conditions change. | The absence of summer activity does not automatically establish absence in spring. |
| Summer | Some features may support breeding groups and dependent young. | Avoid disturbing a sensitive roost while gathering evidence or planning work. |
| Autumn | Mating, roost switching and preparation for winter can affect use. | Autumn is not a blanket permission to remove potential features. |
| Winter | Some species hibernate in trees; activity outside the roost may be limited. | A quiet night or leaf-off inspection cannot alone rule out winter use. |
Leaf-off conditions can improve the view of a tree’s structure. However, better visibility does not make every cavity accessible or reveal all occupants. The ecologist should distinguish the best time to map features from the best time to answer a question about their use.
The same tree can present different survey constraints across the year. Leaf cover may hide a high opening; poor weather may reduce activity; an unsafe branch may prevent close access. A programme should account for these limitations instead of treating a calendar month as a guarantee.
If the design changes or work is delayed, ask whether the earlier findings remain applicable. The relevant question is whether the evidence still supports the current operation.
Use our bat survey season guide for wider survey timing, then agree the tree-specific programme.
Read the evidence: BCT: woodland specialists · BCT survey guidelines, fourth edition (2023; amended March 2024).
12 · Before work is instructed
A recommendation to prune or fell a tree answers an arboricultural question. It does not automatically resolve protected-species requirements. The ecologist needs the actual proposed operation, not just a note that says “tree works”.
For example, crown reduction, removal of a damaged limb and complete felling can affect different features. Access machinery, temporary lighting and the treatment of retained sections also need to be considered where relevant.
Where a tree presents an immediate safety concern, keep people away from the affected area and obtain urgent professional advice. The arborist and ecologist should coordinate the risk response and any necessary authority involvement. Do not assume that a safety concern removes every legal requirement.
Our arborist–ecologist handover checklist covers document coordination. This guide supplies the ecological background behind that handover.
13 · Protection and permissions
In England, bats and their breeding and resting places are protected. Activities that damage a roost, obstruct access or unlawfully disturb bats can create an offence. Protection is not limited to the nights when bats are seen leaving.
Planning permission, tree consent and a bat mitigation licence serve different purposes. One does not automatically replace the others. Where impacts cannot be avoided, the ecologist must assess the appropriate licensing route and evidence needed before the relevant work.
This legal context follows Natural England’s guidance for England. Wales, Scotland and Northern Ireland have their own licensing authorities and procedures. The project location therefore matters.
14 · Long-term stewardship
Old trees and standing deadwood can provide valuable habitat. At the same time, a damaged tree may present a real safety issue. Neither value nor risk should be inferred from appearance alone; competent arboricultural and ecological advice needs to be brought together.
Disease and storm damage can change the available roost resource. Map the work, record decisions and consider how a sequence of operations affects the wider woodland. Routine forestry and development projects may follow different consenting and assessment routes.
BCT’s current woodland resource brings together guidance on ash dieback, forestry operations and protected species. It also signposts the 2025 Bats and Woodland publication. Use the relevant national guidance and keep records of the information used to reach decisions.
Read the evidence: BCT: woodland management and ash dieback guidance.
15 · Compensation is a design decision
A box may provide an additional opportunity or form part of a justified compensation design. However, installing boxes does not itself authorise the loss of a protected roost. Suitability, location, timing and the affected species all matter.
An established cavity has a particular microclimate, entrance and landscape setting. A manufactured box may behave differently and may not be occupied. Compensation should therefore explain the ecological function it aims to provide, how it will be secured and how success will be assessed.
Retention and impact avoidance come first. Where an agreed design includes compensation, consider maintenance, access, overheating risk and monitoring alongside the installation itself.
Explore our bat mitigation and roost design guide and post-development monitoring guidance for the next stages.
Read the evidence: Natural England: bats and planning decisions.
16 · From findings to decisions
A useful report makes the next decision easier. It should explain why more evidence is needed, or why a proportionate avoidance or management approach is appropriate. If the layout, lighting or tree works change, check whether the conclusions still apply.
17 · Practical support
Send us the tree survey or schedule, location plan, ecological reports and a clear description of the intended works. If there are photographs, link them to the tree identifiers and take them only from a safe position.
We can review the ecological question alongside your proposal, explain the evidence route and discuss options for avoiding impacts. Where specialist access, additional surveys or licensing input is needed, the scope and responsibilities should be agreed clearly.
Our starting point is not a predetermined package of extra visits. It is understanding what you want to do, what may be affected and whether a better design can resolve the issue.
For the wider service, see bat surveys. For the arboricultural scope, see tree surveys.
18 · Compact answers
Bats roost in existing cavities and crevices. They do not build nests or excavate new holes. A woodpecker hole, split or sheltered bark feature may provide suitable space.
Tree size alone cannot rule a roost out. The presence of a suitable feature, its context and the evidence matter. Older trees often provide more opportunities, but an assessment should not rely on age alone.
No. A hole may be unsuitable or unused. The assessment must separate potential suitability from evidence of occupation, and then consider the effect of the proposed work.
An ecologist assesses the tree and its visible potential roost features from the ground. The record explains evidence and limitations. However, this differs from a tree condition survey or a closer inspection of a cavity.
FAR is a tree-level category indicating that further assessment is required to establish whether features are present. PRF-M describes a particular feature suitable for multiple bats, potentially including a maternity group. They answer different questions.
No. It is a suitability category, not a work authorisation. An ecologist must consider the proposal, context and an appropriate evidence or precautionary approach, including any legal requirements.
Do not remove or disturb material to look for bats. It may obscure or form part of a roost feature. Ask an appropriately experienced ecologist to assess the tree and the proposed work.
No. A detector samples sound at its location. Quiet bats, hidden exits and intermittent use can be missed. Passing activity also does not prove that bats roost in that tree.
Possibly, but winter is not a universal safe period. Some bats hibernate in trees. The appropriate timing and method depend on the feature, species, evidence and proposal.
Yes, where it genuinely avoids relevant impacts. Retaining the tree alone may not be enough if lighting, access or habitat connections change. The ecologist should record why the revised proposal supports a different evidence requirement.
Not automatically. A box must serve a defined ecological purpose within an agreed design. It does not replace a licence where one is needed or guarantee that the lost roost function will be recreated.
Keep people away from the affected area and seek urgent arboricultural and ecological advice. The response should address safety and protected species together. Do not use a general website guide as permission for a particular operation.
19 · Sources, scope and limitations
Reviewed on 15 November 2026. Professional guidance informs the survey framework; research helps explain ecological behaviour. Neither removes the need for a site-specific judgement.
For the feature-identification literature, Bat Roosts in Trees (Bat Tree Habitat Key, 2018) is specialist further reading. It offers a route into the specialist tree-roost literature.
The illustrations are original educational artwork. They are not survey photographs, measured tree sections or evidence from an ACP project.
Bring the design and ecology together
Share the proposal early so the next step addresses the real impact and the information you already have.