NatureScot Research Report 1368 - How to create woodlands that are resilient in the presence of beaver (Castor fiber): a review of current evidence

Published: 2025

Authors: Kelsey Wilson (Independent Consultant), Jenny Bryce (NatureScot), Alana Skilbeck, Rob Needham, Sheelagh McAllister and Róisín Campbell-Palmer (Beaver Trust)

Cite as: Wilson, K., Bryce, J., Skilbeck, A., Needham, R., McAllister, S., and Campbell-Palmer, R. 2025. How to create woodlands that are resilient in the presence of beaver (Castor fiber): a review of current evidence. NatureScot Research Report 1368.

1. Introduction

Woodland expansion is important for mitigating the effects of climate change and addressing the biodiversity crisis. Riparian woodlands are typically described as those within 50 metres (m) either side of a watercourse and are particularly valuable habitats providing numerous ecosystem services (Cole et al., 2020). For example, riparian woodlands can act as biodiversity hotspots, prevent riverbank erosion, and serve as wildlife corridors (Corbacho et al., 2003; Rood et al., 2015; Bateman & Merritt, 2020). They can also play a vital role in buffering the effects of agricultural diffuse pollution on aquatic ecosystems (Turunen et al., 2019). Additionally, riparian woodlands can contribute to counteracting the effects of climate change by easing rising water temperatures through increased shading (Justice et al., 2017; Turunen et al., 2021). 

A study of Scotland’s riparian woodland network demonstrated that 56% was classed as ‘poor condition’ in 2015-2016 (Ogilvy et al., 2022). A more recent analysis, focusing on areas within 20 m of watercourses, found only 19% of riverbanks in Scotland are wooded (NatureScot, 2023). In July 2023, Scottish Forestry launched the ‘Woodlands for Riparian Benefits Forestry Grant Scheme,’ targeting specific areas to enhance Scotland’s riparian woodlands. Reflecting the expected benefits and the current status of riparian woodland in Scotland, the scheme offers increased grant rates for areas along watercourses. Around 175,000 hectares (ha) have been identified where planting riparian woodland can be prioritised to deliver and maximise multiple benefits. 

One of the main challenges to woodland expansion and condition in Britain is the high-intensity browsing by red (Cervus elaphus) and roe deer (Capreolus capreolus) (Gill, 1992; Gill & Morgan, 2010; Reimoser & Putman, 2011; Tanentzap et al., 2013; Ramirez et al., 2018; Spake et al., 2020). In recent years, an additional herbivore has re-emerged in British riparian woodlands. As Eurasian beaver (Castor fiber) distribution continues to expand throughout Britain, both from natural colonisation and through licenced translocations, the overlap between planned or existing woodland creation schemes and active beaver territories is inevitable. Despite this, there is a notable lack of information available for land managers when designing new woodland creation schemes and managing existing schemes in the presence of beavers.

Beaver can have profound effects on riparian woodlands through browsing and damming (Rosell et al., 2005). For example, the felling of tall trees creates localised, well-lit canopy gaps which can promote tree seedling recruitment and woodland regeneration (Nummi & Kuuluvainen, 2013). Beaver browsing can stimulate regenerative responses in trees in the form of secondary shoots that grow from around the cut stems, as well as increasing structural heterogeneity (Jones et al., 2009; Wilson, 2022) which is associated with increased biodiversity value (Fuentes-Montemayor et al., 2013; 2020). However, some beaver impacts can lead to human-beaver conflict, particularly in intensively managed landscapes (Wróbel, 2020). For instance, localised flooding of commercial woodlands through beaver damming can lead to tree deaths (Hyvönen & Nummi, 2008). Beavers can also uproot newly planted seedlings (Reeves et al., 2016) or fell valuable ornamental trees (Campbell-Palmer et al., 2016). 

While deer browsing impacts can extend throughout entire plantations (Gill, 1992), beaver impacts are typically confined within 30 m of the watercourse (Haarberg & Rosell, 2006). Both beaver and deer seasonally shift their diets with a focus on woody species in the winter and herbaceous vegetation in summer (Spitzer et al., 2020; Mikulka et al. 2022b). However, even in sparsely forested lowland agricultural landscapes, beavers are dependent on a sufficient supply of woody plants to survive the winter (Mikulka et al. 2022c). 

An earlier review of beaver impacts, drawing on experience from seven European countries, found that most woodland managers viewed beaver impacts as positive or a nuisance rather than economically significant (Reynolds, 2000). Felling impacts are described as not being of an economic scale, with dam induced flooding affecting relatively limited areas of forest (typically <1%) and occurring in areas of lower productivity or natural wetlands. Damage to fruit trees was highlighted in France and Switzerland. In riparian woodlands, the effects are characterised by thinning of trees and coppice shrubs rather than clear felling, with the potential for localised depletion of resources, where these are sub-optimal or fragmented (Reynolds, 2000). 

This report seeks to collate information to inform guidance for designing and protecting woodland creation schemes in the presence of beavers. The Beavers in Scotland Report (Gaywood et al., 2015) provides a summary of current knowledge of beaver effects on woodland and species of conservation interest; hence this report does not seek to replicate this information. However, management methods will also have a bearing on other habitats and species, and these are briefly acknowledged. The scope of the report is ‘woodland creation’ in the presence of beaver. Beaver impacts are expected to be focussed within the aquatic and riparian zones (Figure 1). The emphasis of this report is on woodland creation rather than woodland management, though some of the findings will also be relevant to the management of existing woodlands. The scientific literature features more information on impacts on planted woodlands rather than naturally regenerating woods, possibly reflecting forestry practice, however, the implications of beaver presence are relevant to both naturally regenerating and planted woodlands.

1.1 Study aims

The aim of this study is to make recommendations for designing and protecting woodland creation schemes in the presence of beavers – it is not in itself a guidance note. This study is structured around three main components:

1. A literature search and review of forestry practice in other European countries and in North America in relation to woodland creation strategies in the presence of beavers
2. A questionnaire of European personnel involved in woodland management and beaver mitigation to assess the role of beavers (and other herbivores) in woodland creation.
3. Field surveys of woodland creation sites in Scotland to investigate the current nature and scale of beaver impacts.

2. Methods

2.1 Literature review

A literature search using Google Scholar and Web of Science was conducted to find all English-written data-derived, peer-reviewed and ‘grey’ literature that explicitly refers to woodland management in the presence of beavers. Establishing general forestry practice in other countries was outside the scope of this study. Limiting the search to English language documents may have limited access to other useful material but was also outwith the scope of the project. 

The search terms used included ‘beaver,’ ‘forest/woodland,’ and the following: ‘management’, ‘creation’, ‘design’, ‘planting’ or ‘regeneration’. The ‘forest’ and ‘woodland’ search terms were used interchangeably. A database was then compiled with the authors, article title, year, journal, and location of the study. Articles citing retrieved literature and references within were thoroughly searched for additional, relevant material and incorporated into the database (available as a supplementary electronic spreadsheet). 

2.2 Questionnaire

The questionnaire consisted of 20 questions and ‘Google Forms’ was used to facilitate the collection of responses. Invitations to the questionnaire were distributed by email to land managers who have known experience working in the forestry sector. The questionnaire was sent to 10 contacts based in Germany, England, Norway, Scotland, Denmark and Wales who were known to the authors from attendance at European research conferences and meetings. They were invited to pass the questionnaire on to other relevant contacts and hence the final number of recipients is unknown. Whilst recipients were sourced from known contacts, they are considered to be representative/ knowledgeable of their sector. The questionnaire remained accessible to participants for around eight weeks between June and July 2023. The full list of questions and a summary of responses is listed in Appendix C. 

2.3 Site visits

2.3.1 Study sites

Site visits of 14 riparian woodland creation schemes were conducted to assess the extent of current beaver impacts in Scottish woodlands in areas with beavers. The study sites were identified through a preliminary GIS analysis conducted by Scottish Forestry/NatureScot which identified 13 woodland creation schemes that overlap with known beaver territories from the latest beaver census by Campbell-Palmer et al. (2021). Woodland creation schemes were mainly established through Rural Development Contracts (RDC) (2007-2014) and Forestry Grant Schemes (FGS) (2015 to date). Some beaver territories overlapped with more than one woodland creation scheme. One further (privately funded) plantation site was identified post-hoc via the land manager contacting NatureScot for advice. In total, 14 woodland creation schemes across 10 beaver territories in the Tay and Forth catchments were surveyed (Figure 2). All involved planting rather than natural regeneration.

For each scheme, general woodland information was obtained from records and woodland managers where possible. This included planted area (ha), stocking density in trees per ha (tph) and general woodland objectives (e.g., carbon sequestration, native woodland planting. 

2.3.2 Field surveys

Field surveys were conducted between May and August 2023 (by KW, RCP, and SM. During site visits, each woodland creation scheme was first assessed for evidence of beaver browsing using methods by Campbell-Palmer et al. (2021). This involved a thorough walk-through of the riparian zone (50 m from the watercourse) while scanning for any evidence of beaver activity, particularly browsing. Potential entry points were examined by checking fence lines for evidence of beaver digging, burrowing or obvious openings. The presence of tree protection was noted (i.e., deer fence, stock fence, individual tree tubes) and site photographs were taken. Flooding impacts were assessed based on observations of water impoundment or signs of previous impoundment from the vegetation or from reported impacts.

Where beaver impacts were identified, further surveys (outlined below) were then conducted to assess the extent of beaver browsing and flooding. Beaver-impacted woodland creation sites are presented as individual case studies.

2.3.2.1 Beaver impacts

Often it is not practical to examine every tree in a plantation and damage assessment methods typically revolve around examining a representative sample. The method that is typically used for woodland (deer) is the ‘Nearest Neighbour’ (NN) method outlined in Pepper (1998). It involves the selection of a number of points (cluster points) evenly spread throughout the area. Around each cluster point, a predetermined number of trees are assessed for damage. Trees in a cluster are chosen objectively and independently of the damage that has occurred. This method was therefore employed but adapted so that all clusters were within the riparian zone (i.e., 50 m of the watercourse) to include the focal area of beaver activity. As each woodland compartment differed in size, cluster points were calculated for each compartment to adequately cover the riparian zone. Typically, this resulted in two parallel lines located at 10 m and 20 m from the watercourse with clusters placed 10 m apart (Figure 3).

A total of 20 clusters were established in each woodland and five trees were surveyed at each cluster. This resulted in a total of 100 trees being assessed in each woodland creation scheme that was impacted by beavers. For each tree the species and diameter (cm; recorded at 20 cm from the ground) were noted. Trees were also assigned to one of five categories according to their ‘status’, primarily based on the type or presence/absence of beaver browsing (descriptions in Table 1). In beaver-browsed trees, height of cut (cm) from the ground was also noted. This was measured as the highest part of a remaining stump from the ground (or of any gnawing on any partially cut stems). 

Whilst the NN methodology was used to capture beaver impacts within the riparian zone, additional measurements were taken to assess the impact on the woodland creation scheme as a whole. This involved a systematic walkover of all suitable beaver habitat within the scheme using a clicker counter to record every beaver-browsed tree.

2.3.2.2 Deer impacts

To gain further insight into potential impacts of deer and other herbivores, the WHIA (Woodland Herbivore Impact Assessment)method (Armstrong et al., 2023) was carried out at every site. The method involves stopping at ten ‘typical’ locations within the woodland. Although the NN methods were focused within 50 m of the watercourse, as deer and livestock typically range over a greater area, the WHIA plots were located throughout the entire woodland creation scheme with a view to assessing herbivore levels in the wider area. In practice, all but one off the woodland creation sites was deer fenced, so these assessments did not provide much additional information.

The WHIA method is based on observations of the effects of recent trampling, browsing, or grazing by large herbivores on seven main indicators of ground cover, tree, and other plant indicators (Table 2). The level of impact on each indicator is visually allocated based on one of five categories from ‘no impact’ to ‘very high.’ These are then used to determine an overall impact level. Full methods and descriptions are detailed in Armstrong et al. (2023). There is currently no indicator that looks specifically at deer browsing on regenerating beaver cut shoots. 

2.3.3 Data analyses

2.3.3.1 Beaver impacts

The mean diameter (cm) and height of cut (cm) were calculated for each woodland creation site with beaver impacts. In beaver-impacted sites, beaver-browsing was calculated at two distinct levels: a) riparian zone and b) entire woodland creation scheme. Riparian zone beaver-browsing impacts (%) were derived from the NN methods and involved dividing the total number of browsed trees by the total number of trees surveyed (Figure 4 a).

In order to assess beaver-browsing impacts on the woodland creation scheme as a whole, additional calculations were made. Using the provided information on stocking density and planting areas, we estimated the total number of planted trees in the scheme. This figure was then used to calculate the overall beaver-browsing impact (%) across the entire woodland creation scheme by dividing the number of beaver-browsed stems in scheme by the total extrapolated number of stems (Figure 4). 

Beaver foraging preference for a particular tree species was evaluated separately for each site where possible, using Ivlev’s electivity index (Ivlev, 1962). The index ranges from -1 to +1, where negative values signify avoidance and positive values indicate preference. The electivity index for each tree species was calculated using the formula (E = (B – A) / (B + A)), where B represents the proportion of trees of the focal species in the sample of browsed trees, and A represents the proportion of trees of the focal species in the total sample.

2.3.2.2 GIS data

Additional data for each woodland creation scheme was extrapolated from pre-existing Geographic Information Systems (GIS) datasets where possible. The Beaver Forage Index (BFI) model (Graham et al., 2020) was overlaid with the woodland creation boundaries to explore potential relationships with impacts. The BFI shows the suitability of beaver forage within 100 m of all watercourses within Scotland based on a scale of 0 to 5 (Table 3); noting this may change over time where there is a woodland creation objective. The mean BFI was calculated for each woodland creation to assess the suitability of forage within scheme boundaries. Mean BFI was also calculated within the respective beaver territory as a measure of assessing the suitability of surrounding (alternative) resources. 

The latest beaver census GIS data (Campbell-Palmer et al., 2021) was overlaid with woodland creation boundaries to calculate approximate distance (m) from the nearest known beaver lodge or primary burrow (if identified). This could not be accurately determined for every woodland creation scheme because not every beaver territory in the census had a mapped lodge/burrow.

3. Results

3.1. Literature review

The literature search did not reveal any articles that specifically formulates guidance for woodland creation or management in the presence of beavers. There were 36 articles documenting some aspect of woodland design, protection, or management in the presence of beavers, published between 2003 and 2024. Of these articles, 12 were based in North America and 24 in Europe. Stringer & Gaywood (2016) suggest the impacts of the species are broadly similar and reported differences relate to the different habitats in which they have been studied. Hence the findings of studies on either species are taken together and organised and discussed below under headings based on the main management implications. 

One article considered to be of particular relevance is Mikulka et al. (2022a) from the Czech Republic which describes the use of beaver-favoured woody species and naturalised bankside vegetation as a protective measure for commercial stands and natural woodlands. A second article of note, the Planning for Beavers Manual (King County, 2022) describes a proactive approach to catchment restoration and planning. A third article, Piton et al. (2020), assesses the implications of beaver browsing preferences for riverbank (green) bioengineering. An earlier review provides additional insight to the questionnaire on the likely significance of beaver impacts on woodlands in Europe (Reynolds, 2000).

3.1.1 Establishing woodland objectives

Many woodland creation schemes are established to complement current land use practices such as farming and sporting (Sing et al., 2013). The majority of woodland creation in the UK in the last century has been dominated by non-native coniferous plantations primarily for timber production. However, there has been a gradual shift towards planting more native woodlands with multiple objectives (Harmer et al., 2015). Typical objectives for woodland creation schemes can include planting woodlands for biodiversity, flood mitigation, carbon sequestration, or commercially productive woodland (Scottish Forestry, 2020). The objectives will have a significant bearing on the approach to woodland creation, whether by natural regeneration or planting, such as tree species choice and design. For example, the Keeping Rivers Cool Manual (Lenane, 2016) explores the characteristics of riparian woodland that will influence water temperatures, principally for enhancing rivers for fish. The manual suggests the length of continuous tree cover, width of riparian tree cover, tree density and amount of shade cast are important factors. 

Beaver browsing and damming are a natural part of riparian woodland ecology (Westbrook, 2021). Beaver activity is known to enhance species richness (Stringer & Gaywood, 2016) and improve river flow attenuation (Puttock et al., 2017, 2021). Therefore, certain woodland objectives, such as managing for biodiversity and flood mitigation, could be met (and even enhanced) in the presence of beavers. Although conifers are not the preferred species for beavers, they can still be impacted with incidences of ring barking and flooding observed (Reynolds, 2000; Iason et al., 2014). Scottish Forestry has supported the planting of almost 3,000 ha of broadleaved woodland for quality timber production under the Forestry Grant Scheme and many new native woodlands will also have the potential for firewood production (Scottish Forestry, pers comm). In the last decade, beaver impacts on commercial forests in the Czech Republic have resulted in costs of over € 2.5 million (Mikulka et al., 2022a). Hence, considering what might be realistic objectives for the riparian or wetland components of a woodland creation/ regeneration scheme where beavers are present or are likely to become present in future years is advised. Woodland is considered most vulnerable to beaver browsing in the first 20 years of establishment (Mikulka et al. 2022a).

Determining at an early stage the extent to which beavers can be accommodated will also be important. In a guidance manual for habitat restoration projects in the presence of the North American beaver (Castor canadensis), it is considered essential to plan for the ‘level of beaver integration’ (King County, 2022). A proactive approach is likely to provide a more robust woodland design and protection strategy that aligns with woodland objectives. Beaver integration options could be characterised as: a) no allowance for beavers to be present or have impacts, b) accepting low-level beaver impacts and mitigating accordingly and c) allowing beavers to be an integral part of the woodland design (see Table 4). Typically, this decision process is already undertaken with regard to deer (Scottish Forestry, 2020; Woodland Trust, 2022).

3.1.2 Managing herbivore impacts

When young trees are regenerating or newly planted, they are particularly vulnerable to herbivory because they have limited energy reserves and root systems to support their growth and development. However, as trees mature and develop a more extensive root system, they generally become better equipped to withstand browsing pressure (Crawley, 1984), although the ability to repeatedly sprout from cut shoots varies between tree species. It is well-known that excessive browsing impacts from deer can have considerable constraints on tree establishment (Gill, 1992; Ramirez et al., 2018). Woodland regeneration or planting schemes throughout Britain are therefore typically protected from deer (Cervidae), hare (Lepus europaeus), rabbit (Oryctolagus cuniculus), squirrel (Sciurus spp.), and vole (Microtus spp.) browsing via fencing, individual tree guards, animal culling, or some combination of all three (Redick & Jacobs, 2020). The comparative costs of individual tree protection, fencing and deer management are scale dependent (Hodge & Pepper, 1998). Herbivore impact management approaches are therefore an important consideration, not just in the presence (or future presence) of beavers, but with regard to other herbivores.

3.1.2.1 Woodland-level protection

Whilst many woodland creation schemes in Scotland have traditionally relied on deer fencing there is growing recognition of the costs and environmental impact of fencing, with calls for greater focus on deer management (DEC, 2004; Pepper et al., 2020; NatureScot, 2022). The length of fence line required to protect purely riparian schemes means that this approach is unlikely to be cost effective. Most woodland creation schemes are much larger than the riparian zone, with fencing well back from the river. Hence where fencing is considered to be appropriate, fence lines and design would need to be carefully considered with multiple herbivores in mind. 

Exclusion fence specifications (for any species) are highly site-specific. Effective designs are likely to vary depending on topography, substrate, and distance from the watercourse. Deer fences are usually 1.8 m in height and can have a netted lower section to further exclude rabbits and hares. Mesh sizes used for deer fencing commonly range from 15 x 20 centimetres (cm) up to 30 x 22 cm (Trout & Pepper, 2006). To exclude beavers, mesh size should not exceed 10 x 10 cm in order to prevent beavers entering and/or becoming trapped (Vorel et al., 2016). Beaver fatalities have been documented as a result of attempts to fit through larger wire mesh gaps in fencing (Rosell & Campbell-Palmer, 2022).

Deer fences could be adapted to create a dual-purpose fence that also excludes beavers through the addition of a ‘mesh-skirt’ to prevent beavers gaining access by digging under the fence line. This modification involves the addition of a bent bottom section of wire mesh that forms an L-shape flush to the ground (Campbell-Palmer et al., 2016; King County, 2022). Fences can also be dug partially underground to prevent beaver access, as demonstrated by Kamczyc et al. (2016) as an effective means of protecting riparian oak (Quercus spp.) stands. Their fencing was situated parallel to the water and extended up either side of the plantation by 40 m. Similar guidance has been issued as a preventative measure for tree protection along the Danube River Basin (Valachovic, 2014) and in the Czech Republic (Vorel et al., 2016). 

In scenarios where deer-fencing is specifically not necessary/ appropriate then 1.2 m stock fence with mesh-skirting could be used (see Figure 5). In North America, Nolte (2003) demonstrated that 0.95 m high fencing can effectively reduce beaver browsing impacts in small, targeted areas. Mesh skirting could potentially be added onto existing fences in woodland creation schemes that are facing beaver impact issues. Detailed beaver fencing specifications are described in Appendix 1.

Electric fences have also been suggested to be effective in deterring beavers, but is considered more viable as a short-term measure, for example, where there is damage to high value crops. The use of electric fences for beaver also has the potential to create concerns for animal welfare (Campbell-Palmer et al., 2016). 

While deer fencing can prove effective for assisting woodland establishment, the following limitations are likely to further limit the application of beaver exclusion fencing for woodland creation in Scotland:

1. Beaver fencing is financially costly and likely to exceed the cost of deer fencing. The cost of fencing is likely to be too expensive/ impractical on smaller purely riparian schemes for grant support.
2. Fencing faces the risk of being damaged by the force of rivers and debris during flooding events, requiring ongoing maintenance and repair (King County, 2022). Hence fencing is often considered not to be appropriate near to ‘flashy’ rivers for this reason.
3. Beaver exclusion is also likely to be challenging where there are watercourses within the woodland as the designs would need to incorporate ‘beaver proof’ water gates.
4. As for deer fencing, there are a range of other environmental costs; fragmentation of habitats and natural movement patterns (Vorel et al., 2016) e.g. consideration of impacts on Eurasian otter (Lutra Lutra), landscape impacts, bird collisions, impacts on lichen and bryophyte communities through affecting stand density and shade regimes. 

A suggested alternative involves creating 'odour fences' by establishing paths regularly used by humans and their dogs (Loeb et al., 2014). However, with Scottish beaver territories already thriving in semi-urban areas featuring busy dog-walking paths, fishing beats and other recreational areas (Campbell-Palmer et al., 2021), beavers appear to be relatively unaffected by human disturbance. Therefore, the practical application of this approach is likely to be limited.

3.1.2.2 Individual tree protection

For smaller riparian schemes, individual tree protection may be a more cost-effective option e.g. protecting existing valuable mature or ornamental trees. Many woodland creation schemes protect trees from herbivore browsing with individual tree guards (plastic tubes 0.6 - 1.8 m tall) or vole guards (0.2 m tall) (Scottish Forestry, 2020; Woodland Trust, 2022). However, reports of beavers gnawing through plastic tree tubes have been well documented (Figure 6).

To safeguard individual trees from beaver browsing, trees can be wrapped with wire mesh. A report detailing guidance for coexisting with beavers in the Czech Republic emphasises the importance of wrapping buttress roots and low-hanging limbs without hindering tree growth (Campbell-Palmer et al., 2016; Vorel et al., 2016). A recent study conducted in North America trialled wire-mesh tree guards and found them to be effective against beaver browsing when installed properly (Westbrook & England, 2022). Detailed methods for tree-wrapping are accessible to landowners online through NatureScot (NatureScot, 2020) and are summarised in Appendix B of this study. Caution should be exercised, as many examples of poorly applied and ineffective tree-wrapping have been documented (Figure 7 a, b) (Rosell & Campbell-Palmer, 2022). 

In Scotland, experience of tree protection from beaver impacts is generally with regard to established trees rather than new plantings. There have been limited trials of the protection of new planting using mesh guards (Figure 8). Hall et al. (2011, 2015) report that protection with 1.8 m vented plastic tree guards resulted in a higher proportion of tree survival than trees with a wire mesh guard, most likely due to the difference in microclimate/ competition effects. However, the addition of a wire mesh guard to shorter plastic guards did reduce the amount of browsing on terminal buds. Hence it may be that a combination of protection methods or a novel ‘gnaw’ resistant guard is most effective.

Similarly, the use of anti-game paints on trees has been shown to dissuade beavers from gnawing on them (see Appendix C for details). Tree painting can be especially useful when some trees are too difficult to wrap effectively due to their size or shape (Campbell-Palmer et al., 2016). However, deterrent paint may be less effective on trees of smaller diameter, possibly as beavers can fell these in a few bites without having to endure the gritty texture that they find unpleasant and because the rapid expansion in sapling girth is likely to limit the repellent’s effectiveness and longevity. Combined with the high cost and labour intensiveness of individually painting trees, this approach is unlikely to be applicable for woodland creation but does have a role in protecting high value trees.

An intermediate option between individual and woodland level protection would be the use of small-stand or coupe protection using fencing. This approach is likely to be most applicable where there are groups of trees that are of particular importance or sensitivity, e.g. stands of mature aspen (Populus tremula) in the riparian zone or stands of hazel (Corylus avellana) or willow (Salix spp.) that host rare lichen assemblages.

3.1.2.3 Management of the riparian zone

Managing the riparian woodland specifically for beaver has been suggested by several authors as a potential tool for protecting the adjacent woodlands from beaver browsing impacts (Valachovic, 2014; Gaywood et al., 2015; Rosell & Campbell-Palmer, 2022. Developing a naturalised riparian zone would also provide additional benefits such as increased biodiversity, bank stability, wider habitat connectivity and water temperature regulation through shading (Cole et al., 2020). This idea is further expanded in the next section on woodland design.

3.1.3 Woodland creation design 

3.1.3.1 Species selection

An understanding of beaver browsing preferences is valuable when designing woodland creation schemes. Beavers have been termed as ‘choosy opportunistic generalists’ due to their highly adaptive and selective foraging choices that are dependent on the site-specific availability and composition of the woodland (Vorel et al., 2015). Stem diameters of <10 cm within 10 m of the watercourse are typically preferred (Misiukiewicz et al., 2016; Wazna et al., 2018; Jackowiak et al., 2020). Beavers are thought to display a strong preference for all native willow species¹ and aspen but will regularly select hazel, birch (both Betula pubescens and pendula), and rowan (Sorbus aucuparia) (Iason et al., 2014, Janiszewski et al., 2017; Mikulka et al., 2022b; Wilson, 2022). Although conifers are not typically browsed due to high levels of plant metabolites (Johnston, 2017), beavers are known to occasionally take saplings or ring-bark mature conifers (Iason et al., 2014; Reynolds 2000). In Norway, beavers were found to forage on a minimum of 29 tree species (Haarberg & Rosell, 2006) and may therefore utilise a wide range of tree species in riparian woodlands for foraging and construction (Table 5). Selective foraging by beavers could thereby affect the species composition of woodland (as with other herbivores) leading to more conifer-dominated forests (Pejstrup et al. 2023) and potential loss of rare tree species (Nolet et al. 1994). 

¹ Bay willow (Salix pentandra), crack willow (Salix fragilis), goat willow (Salix caprea), grey willow (Salix cinerea), eared willow, (Salix aurita), common osier (Salix viminalis), white willow (Salix alba) and purple willow (Salix purpurea).

¹ Tree species of wet woodland communities (National Vegetation Classification W1 – W7) (Hall et al., 2004; Forest Research, 2024) recommended for woodland creations (Staton et al., 2024; Woodland Trust, 2022).

² Based on studies in Scotland (Iason et al., 2014; Jones et al., 2009) and tree growth rates/resprouting abilities (Staton et al., 2024). 

³ All of these trees have been documented in beaver diets (i.e., are palatable). Levels of palatability estimated from foraging preferences in studies by Haarberg & Rosell (2006) and Iason et al. (2014). Note that beaver foraging preferences are strongly dictated by the availability of other plant species in the area and may vary.

⁴ From Glenz et al. (2006) and tree flooding tolerances in Staton et al., (2024).

The foraging preferences of beaver present two main strategies for reducing the impact of browsing on woodland creation objectives; the first being to plant unpalatable tree species in order to deter beavers. For example, Piton et al. (2020) propose the use of unpalatable species to minimise browsing impacts of beaver on riverbank bioengineering works. However, this approach seems unlikely to be widely successful for several reasons. Firstly, there are relatively few tree species typically planted that beavers will avoid (Haarberg & Rosell, 2006; Woodland Trust, 2022). Secondly, unpalatable species may not be well adapted to wetter conditions, which could result in higher losses overall. For example, conifers are less tolerant of periodic inundation than broadleaves (Hyvönen & Nummi, 2008). However, alder (Alnus glutinosa) is highly adapted to withstand flooding and is not typically a preferred food source of beavers (unless it is readily available and/or there are few alternative species) (Haarberg & Rosell, 2006; Glenz et al., 2006). Hence alder may provide a helpful component of riparian woodland; noting it is being affected by Phytophthora alni on many Scottish river systems.  

The alternative strategy for tree species selection would involve selecting and over-planting tree species that beavers find palatable (Pollock et al., 2015; Rosell & Campbell-Palmer, 2022; King County, 2022; Mikulka et al. 2022a, 2022b). Coincidentally, species that beavers find palatable are usually well-adapted to withstand periodic flooding and are likely to resprout after beaver browsing (Iason et al., 2014) resulting in a more resilient woodland. Details of the tolerance of tree species to inundation can be found in (FOWARA, 2006 and Glenz et al. 2006). The tolerances cited in FORWARA, 2006 are based on flood heights and the length of inundation. However, most species will experience reduced growth or mortality when the roots are permanently submerged due to anaerobic conditions (with willows recorded as surviving for up to 5 years, as cited in Jones et al. 2009) as may be the case for trees where water levels permanently rise due to beaver dams. Hence topographic areas liable to become wetlands could be deliberately left as open areas. In practice, the riparian zone would be part of the wider woodland creation scheme, but it could have separate woodland objectives and beaver integration goals to the rest of the scheme.

Diversionary food provision has been shown to be an effective mitigation method for reducing impacts of other mammals on forestry stands in species such as red deer (Arnold et al., 2018) and black bear (Ursus americanus) (Ziegltrum, 2004). Researchers in the Czech Republic found that beaver impacts on commercial forest stands were significantly reduced by growing densely planted ‘softwood’ (note this term is used to refer to willows or poplar (Populus spp., including aspen), rather than conifers as the term is used in the UK) at a distance of 10-20 m from the watercourse (Mikulka et al., 2022a). The authors also propose that beaver impacts can be significantly reduced by promoting a more naturalised bankside vegetation. They make recommendations for willow/ aspen planting based on distance from the water, the commercial tree species being grown and tree diameter. Guidance for the Danube River Basin also states that a 10 m wide strip of shrub willows can provide alternative food sources and lower damage to more valuable trees (Valachovic, 2014). 

In North America, King County (2022) proposed that aspen could be planted strategically in one area to divert browsing from another area. Aspen is highly preferred by Eurasian beavers (Haarberg & Rosell, 2006; Jones et al., 2009) and already features as a recommended planting species in native broadleaf schemes (Woodland Trust, 2022). Beaver have a particular preference for aspen, which is a species of conservation interest in its own right in Scotland and is associated with many specialist species (Parrot & Mackenzie, 2009). The capability of aspen to reproduce by suckering perhaps allows greater resilience to browsing impacts, but some of the specialist species it hosts require particular life-stages of aspen, including mature trees and deadwood from large trees. Hence consideration of increasing aspen as a component of the woodland both within the riparian zone and more distant from the watercourse would be beneficial to restoring this species.

Willow spp. are not typically used in woodland creation schemes in Scotland or only as a minor species due to its shrub-like form (Woodland Trust, 2022). Both aspen and willow are fast-growing, highly regenerative species of biodiversity value that can potentially reach diameters within the lower range of beaver foraging preferences (i.e., 2-3 cm) in just a few growing seasons (Haarberg & Rosell, 2006; Woodland Trust, 2022). Beavers generally select trees greater than 2 cm in diameter (Haarberg & Rosell, 2006), therefore new plantings may not be targeted, but impacts will likely depend on the availability of alternative forage. Hence there is a danger trees may be impacted before they can develop herbivory responses (i.e., regrowth) (Martell et al., 2006). Hall et al. (2015) found greater success of riparian tree establishment in ‘drylands’, implementing deep planting of cut willow stems using a mechanical auger to allow roots to establish directly into the water table. 

Allowing a riparian zone to regenerate naturally may be preferable due to the lower establishment costs and potentially lower nutritional value/palatability of natural regeneration. Natural regeneration will provide less opportunity to manipulate the species composition but is likely to be dominated by native trees and shrubs likely to provide suitable beaver habitat. The suitability of the site for natural regeneration will depend on the objectives of woodland creation, suitable seed sources, and the approach to managing large herbivore impacts.

3.1.3.2 Stocking density 

When designing woodlands in the presence of beavers, selecting an appropriate stocking density could also help to mitigate potential browsing or flooding impacts. Typical planting densities vary between 1,100 - 1,600 tph for broadleaves (Scottish Forestry, 2020). Tree mortality due to herbivory is usually expressed as a percentage of the total scheme (Pepper, 1998). Therefore, from a practical perspective, percentage losses could be minimised if more trees are planted initially, though this would incur higher establishment costs. Nolet (1997) estimated that at a willow density of 0.1 m^-2, an average family of 4.6 beavers with 1.8 ha of wooded banks in their territory, would remove 32% of annual growth, and suggested this level of exploitation is sustainable. A density of 0.1 m^-2 is equivalent to 1000 stemsper ha so would be within the normal range for planting broadleaves in Scotland. 

A case study in North America, found that although beaver browsing and flooding impacts of planted trees were common, tree survival was not unacceptably low because the initial planting strategy incorporated large numbers of trees (Pollock et al., 2015). Guidance for restoration projects in North America promotes the use of higher stocking densities a) because there will be more trees to mitigate losses, but also b) a higher density means there will be more stems present to resprout and regrow (King County, 2022). They further emphasise that although this approach will result in additional upfront costs associated with increased planting efforts, fewer costs will be diverted to beating up (King County, 2022).

Research on mixed-species woodland plots in Knapdale Forest in Scotland showed that the intensity of beaver browsing was not related to tree density (Wilson, 2022). Studies in Canada have shown similar results where tree density has no significant impact on beaver forage selection (Gerwing et al., 2013). Conversely, studies in the Czech Republic found varying relationships with stand density and commercial crop damage depending on the tree species. They suggest beavers selected high density and diversity stands with trees of mixed age and more undergrowth (Mikulka et al. 2022a). 

3.1.3.3 Woodland creation area

The location and extent are likely to be significant factors in the success of a woodland creation scheme in areas with beavers. As central-place foragers, most beaver browsing is recorded within 10 m of the watercourse (Janiszewski et al., 2017; Wazna et al., 2018; Jackowiak et al., 2020). However, beaver damaged trees have been reported up to 100 m from the watercourse (Reynolds, 2000), with distances tending to be greater in more natural forests likely due to the distribution of preferred tree species (Mikulka et al., 2022a). Pejstrup et al. (2023) recorded more beaver felled trees in territories with high beaver activity (>2 lodges and/or dams per kilometre (km) of shoreline). In Knapdale Forest, 90% of all beaver browsing activity in woodland plots was concentrated within 10 m of the shore (Wilson, 2022). Hence beaver impacts are more likely to feature in areas that overlap with the focal area of beaver activity, particularly within 10 m of the watercourse. However, since beavers can modify the landscape through damming and canal digging, the area of focal activity could shift over time.

As beavers are highly territorial, browsing and flooding impacts will be spatially and temporally patchy (Kivinen et al. 2020). Piton et al. (2020) investigated beaver foraging preferences on the bank of a braided alpine river that was being recolonised by woody shrubs (willows and tamarisk sp.) following cutting the previous winter. They found evidence of selection by beaver at two different spatial scales: selection of preferred browsing patches and selection for tree species within the patches. At the site studied, beavers selected four metre wide ‘‘highly-browsed patches’’ that tended to be repeated every 15 m alternating with ‘‘low-browsed patches’’. The patchiness did not reflect tree species distributions and hence was proposed to relate to non-foraging related factors such as distance from the lodge, denser patches providing protection from predators or group foraging. 

Estimates of the daily food requirements of beavers are up to one to two kg of woody vegetation per animal, with one Norwegian study finding between 12 and 47 trees were felled per beaver per month. However, several studies have noted that tree felling typically exceeds the levels necessary for immediate food requirements or building activities, with some felled trees being abandoned (Reynolds, 2000). In employing willow/aspen plantings as a diversionary tactic, Mikulka et al. (2022a) recommend dense stands are planted at a minimum of 0.3 ha per beaver territory or 0.1 ha per km of watercourse, to account for resource depletion. Beavers are considered more likely to locally deplete their food resources where there are preferred species present, or in sub-optimal habitats. For example, in Switzerland small areas of habitat are considered vulnerable to over-exploitation and require restocking with fast-growing willow species (Stocker, 1985 as cited in Reynolds, 2000). 

Parker & Rosell (2003) emphasise that economic losses from beaver are highly dependent upon plantation size. For example, large woodland owners in Norway may typically lose around 0.1% of their conifer production from flooding by beaver over time, which is suggested to be negligible compared to the combined losses from other pressures such as disease or ungulate browsing (Parker & Rosell, 2003). Woodland creation in Scotland may not be feasible at the same scale as central Europe, however, designing (and linking) larger woodland creation schemes could help create a wider network of continuous riparian woodland. A larger woodland resource will undoubtedly help mitigate the impacts of beaver browsing and flooding via a ‘dilution effect’. The corollary is that smaller and more isolated woodlands are likely to experience proportionately higher beaver impacts.

3.1.4 Adaptive beaver monitoring and management

Regular herbivore impact monitoring is necessary for effective woodland management (Thomas et al., 2015; Carpio et al., 2021). In Scotland, guidance recommends that woodland managers assess the extent and impact of browsing herbivores annually (Scottish Forestry, 2020; Forestry Commission, 2021; NatureScot, 2022). Including beavers in herbivore impact assessments would help raise managers’ awareness of beaver presence and to implement adaptive management strategies. In most cases, beaver monitoring efforts could be restricted to within 50 m of the watercourse (Haarberg & Rosell, 2006). Where fencing is employed, this could be as simple as involving a check of fence lines within 50 m of the watercourse for obvious signs of beaver entry coupled with a walkover of the riparian zone for any beaver field signs (as detailed in Campbell-Palmer et al., 2021).

Integrating beaver monitoring into existing herbivore impact assessments would enable beaver-deer impacts to be identified and addressed. Observations in Scotland indicate that deer readily browse the regenerative shoots that sprout from beaver-felled trees (Jones et al., 2009; Iason et al., 2014), but that the degree of impact depends on deer density (Wilson, 2022). One study reported deer browsing on 68% of beaver-browsed trees in riparian woodland plots (Iason et al., 2014). A recent study in Scotland demonstrated that 62% of beaver-felled trees re-sprouted secondary shoots which were concentrated closer to the ground, and richer in nitrogen, than those from standing trees. These findings suggest that beaver browsing could influence deer feeding behaviour and distribution (Wilson et al., 2024). Similar interactions have been found in North America with beaver and elk (Cervus canadensis) (Hood & Bayley, 2008; Loeb & Garner, 2022).

Existing herbivore impact assessment methods do not integrate beaver impacts. Incorporating simple, quantitative measurements of beaver and deer browsing on the secondary shoots resprouting from beaver-felled trees could be one beneficial addition to existing methods. The number of browsed stems could then be compared with the overall stocking density to assess the scale of impact. Changes to these methods would require trialling before they could be widely adopted. The results of monitoring will largely inform how effective deer management is with regard to the woodland scheme either on its own or in combination with beaver browsing as a prompt to review deer management efforts. Acceptable levels of impact should be pre-determined.

A practical example of the benefits of adaptive management is highlighted by Pollock et al. (2015) in North America; offsetting tree losses through replacement planting and adapting the species composition to be more compatible with long-term inundation from flooding by beavers.

3.2 Questionnaire

A total of eight people completed the online questionnaire. This is acknowledged to be a small sample; however, the respondents all have senior ecological/forestry-based roles in the organisations/businesses they represent from across Britain (n=4) and central Europe (Germany, Denmark, and Switzerland) (n=4) and responses provide a clear indication of views. A total of 75% (n=6) of respondents had not encountered forestry practice guidance or recommendations that address the use of beavers as a natural element in woodland creation in their country. The majority of respondents (87.5%, n=7) believe there is insufficient information in current woodland management guidance addressing the impacts of beavers on woodlands, particularly in relation to herbivore management. 

When asked which has the greater influence on woodlands, beaver tree-felling or beaver-induced flooding, the majority of respondents (87.5%, n=7) opted for beaver-induced flooding. When asked what scale of impact they think beaver tree-felling has on woodland, most respondents chose ‘low’ (37.5%, n=3) and ‘medium’ (37.5%, n=3). When asked what scale of impact they think beaver-induced flooding has on woodland the level with most responses was ‘very high’ (37.5%, n=3) (Figure 9).

When asked to choose which has the greater impact on woodland between beaver tree-felling and deer browsing, the majority of respondents chose deer browsing (87.5%, n=7). A total of 62.5% (n=6) of respondents have encountered situations where there were interactions between beavers and other herbivores (e.g., deer) in the woodlands they manage. The remaining respondents said they have not encountered this (25%, n=2) or they do not know enough information to comment (12.5%, n=1).

Participants were prompted to describe some benefits and negatives that beavers can bring to woodland creation and management. Some benefits provided included ‘beavers coppice trees, helping to create healthy and diverse woodland’ and ‘habitats created by beavers are very diverse, dynamic and very species rich’. One drawback was described as ‘if deer numbers are too high, their secondary browsing may prevent coppicing, but this is more a land management issue than a beaver issue’. Other responses stated that ‘beavers could impact newly created woodland by browsing hard on trees before they’ve had time to establish’ and ‘strictly speaking, beavers are a nuisance in productive forestry - mainly because of damming in low-lying areas. The direct impact of felling trees is a minor problem’.

In situations where there were interactions between beavers and other herbivores (e.g., deer) one respondent said they managed these situations by ‘shooting the deer.’ In contrast another respondent said ‘no management is necessary. Deer concentrate browsing at beaver sites, this reduces conflicts in other parts of the forest.’ Some measures that respondents have implemented (or would recommend) to manage the impact of beavers on woodland creation and management include the following: ‘exploring the option of over planting willow for new planting, as well as looking at fencing off newly planted trees’ and ‘creating large willow bushes along the shoreline as preferred and distracting food’. A full list of the responses to all comment questions is detailed in Appendix D.

3.3 Site visits

3.3.1 Woodland creation scheme characteristics 

The 14 woodland creation sites (labelled A-N Table 6.) ranged in size from 0.7 ha up to 81 ha (mean 21.6 ha ±23.7 SD). They were mostly established between 2012 and 2021, with one site originally planted in the 1980s. Species planted at each site mostly comprised a mix of native broadleaves and conifers, with only four woodland creation schemes consisting exclusively of native broadleaves. 

Tree protection from other herbivores varied at each site. Deer fencing, stock fencing and/or individual tree tubes (or some combination) were used. Deer fencing was used at every woodland creation site except for one, which used stock fencing (Table 6). Evidence of woodland herbivore impacts (other than beaver) were largely recorded as ‘no impact’ across most (n=10) of the woodland creation sites. ‘Medium’ impact was noted at three sites and ‘high’ impact was recorded at the site with no deer fencing. 

The main stated objectives for woodland creation consisted of carbon sequestration, flood management, and woodland expansion in line with Scotland’s Forestry Strategy. One site objective was listed as restructuring regeneration i.e., restructuring the tree age and species composition at the point of re-planting following felling.

A total of 12 woodland creation schemes directly overlapped with a previously mapped beaver territory. The remaining two were located within 100 m of a known beaver territory. Distance from the nearest known beaver lodge could be established for eight woodland creation sites and averaged 328 m (±240 SD) and ranged from 20 m – 656 m. Within woodland creation schemes, mean BFI averaged 2.9 (±1.05 SD, range 1.04 - 4.24) suggesting the beaver habitat as ‘moderately suitable’ overall. In beaver territories, mean BFI was similarly estimated as ‘moderately suitable’ and averaged 2.6 (±1.76 SD, range 1.7 - 3.5). 

Table 6. List of 14 woodland creation sites and their area (ha), year of planting or most recent restock, species (NB = native broadleaves, M = mix of native broadleaves and conifers), mean Beaver Forage Index (BFI, scale 1-5), and form of tree protection (DF= deer fence, T = individual tubes, VG = vole guards, SF = stock fence).

*Site planned for restocking in 2024.

3.3.2 Beaver impacts

Beaver browsing was recorded in three of 14 woodland creation schemes (discussed in detail below). Flooding impacts from beaver damming was recorded in one woodland creation site (also impacted by beaver browsing). Beavers had entered two separate compartments of the woodland creation at Site C. For this site, the overall impact was determined as an average (%). 

In the riparian zone of beaver-impacted schemes (within 50 m of watercourse), average beaver browsing was calculated as 38% (range 12 – 54%). As a proportion of total woodland area, beaver browsing impact averaged 4.4% (range 0.1 – 13%). Flooding impacts at the one affected site were historical and therefore the area affected could not be accurately measured (see Case Study A). Beaver entered into two schemes via water gates. Beavers had entered the remaining site by pushing underneath a standard stock fence.

Tree measurements were recorded during NN survey methods on a sub-sample of 100 riparian trees in beaver-impacted woodland creations. Beavers entered two separate compartments in Site C, therefore 400 trees were sampled across three woodland schemes in total. Of the recorded trees, a total of 41% (n=164) were browsed by beaver. The most common type of beaver browsing was ‘b_stump’ in which 81% (n=133) of the browsed trees had been completely felled with only the stump remaining. Lower numbers of browsed trees were recorded as ‘b_p’ (partially-felled; 13%, n=22), ‘b_cut’ (minor branch removed, 3%, n= 5) and ‘b_up’ (gnawing on an upright tree, 2%, n=4).

The diameter of browsed trees varied at an average of 7.2 cm (range 1 – 51 cm, ±7.7 SD) which were cut at an average height of 35.2 cm (range 2 – 67 cm, ±12.6 SD). Although alder was the most commonly selected species across all 3 sites (35%; n=58 of all browsed trees), it was also the most available. Electivity values were highest for willow and aspen (Table 7) indicating a preference for these species. Samples are small in these case studies, but the results are aligned with known preferences.

*A small number of tree stems at Site A (old stumps; n=4) were unidentifiable due to lack of distinguishable features, potentially exacerbated by prolonged flooding.

**Site C had a total of 12 beaver-browsed stems (out of 100 sampled) therefore the electivity values are unlikely to be representative.

3.3.3 Case study A

3.3.3.1 Site description and background

Site A is a small native broadleaf plantation (1.1 ha) which was planted under a SFGS scheme approximately 11 years ago. The plantation was stock-fenced and trees were individually protected with plastic tubes. Stocking density was originally estimated to be 1600 tph. Planted species mainly comprised alder and oak with some smaller numbers of willow, hawthorn (Crataegus monogyna), and ash (Fraxinus excelsior). The plantation occurs in a low-lying area and was established along a section of watercourse which runs adjacent to the entire north stock-fenced boundary of the plantation.

3.3.3.2 Beaver impacts 

During the site visit, an inactive beaver lodge was recorded on the adjacent watercourse approximately 20 m from the plantation boundary. Around six previous beaver entry points into the plantation were recorded along its north boundary (Figure 10). Through NN methods, beaver browsing was estimated to comprise 54% of trees in the riparian zone. Beaver browsing was estimated to total 13% of the entire plantation. 

No evidence of fresh beaver browsing (<6 months old) was recorded, and it was clear that most of the trees had been browsed several years prior. The most common species browsed was oak and the average tree diameter selected was 16.6 cm (±7.7 SD). Trees were felled on average at a height of 44 cm (±9.3 SD). One notable finding included evidence that trees had remained protected until they reached a sufficient size to break from the plastic tree tubes. Beavers had then partially gnawed through the remaining tube to access the tree. 

In addition to browsing impacts, it was evident that the plantation had been historically impacted through flooding. This was known from previous site visits (by RCP) and visually from the plantation’s current appearance (Figure 11). The ground was desiccated with large cracks, with a presence of rushes, and deaths of upstanding (unbrowsed) trees that had experienced prolonged waterlogging. The number or proportion of trees that had been lost or died standing due to waterlogging was not recorded but was notable (Figure 11). The distance of the furthest beaver browsing from the water could not be accurately determined because it would have varied with flooding levels. However, the furthest recorded browsed tree from the existing original watercourse was recorded as 29.4 m. 

3.3.3.3 Deer impacts

The plantation was visibly impacted by deer with a clear browse line developing on several of the remaining oaks. Roe deer tracks were visible in the muddy substrate and had been browsing on the shoots resprouting from beaver-browsed trees. This was the only plantation where deer had been recorded browsing on the regrowth from beaver-browsed trees.

3.3.3.4 Overall factors affecting woodland objectives

Overall, it was estimated that most of the trees in this plantation had died through a) beaver browsing, b) prolonged flooding or c) some combination of both. Woodland objectives for this plantation were thought to be primarily for expansion of native woodland. Given the significant tree losses, it is unlikely that this plantation fulfilled the grant conditions and objectives.

3.3.4 Case study B

3.3.4.1 Site description and background

Site B is located on a sporting estate in Glen Lyon. Two plantations were identified as overlapping with a beaver territory on the south bank of the River Lyon Site B (4.7 ha) and Site D (2.7 ha). Both were planted in 2012 at a stocking density of 1600 tph. Provided planting plans indicated a mix of pine and birch (70%), oak (20%) and smaller proportions of juniper (Juniperus communis) and rowan (10%). Plantations were deer-fenced and trees were also individually protected with plastic tubes. The desired objectives for the plantations are for carbon sequestration and the woodland was funded under the RDC scheme. There were no recorded beaver impacts in Site D and hence the plantations were assessed separately.

3.3.4.2 Beaver impacts site B

During surveys, beaver browsing was recorded in Site B. A single beaver entry point was recorded via a wooden water gate that was situated directly on the main river (Figure 12). Beaver impacts were minimal in this woodland creation scheme and consisted of a single, localised cluster of browsing around the water gate (Figure 13). No impacts of flooding were observed. 

NN methods estimated that 9% of the riparian zone had been impacted by beaver browsing. In reality, a total of 12 trees in the entire 4.7 ha plantation had been browsed by beavers (estimated 0.2% of trees). Beavers had browsed 10 stems of (pre-existing) alder in addition to two (planted) birch stems. The average diameter of browsed trees was 3.6 cm (±1 cm SD). Trees were felled at a height of 43.8 cm (±10 cm SD) from the ground. All beaver-browsed trees were located no further than 7.4 m from the watercourse.

3.3.4.3 Deer impacts

Robust deer fencing had successfully deterred any deer impacts in this plantation and the overall browsing impact was recorded as ‘low’. Some signs of historic cattle trampling and some minor browsing by hare were observed throughout the plantation.

3.3.4.4 Overall factors affecting woodland objectives

At the time of survey, beaver browsing was minimal and concentrated around the single water gate entry. Beavers can still access the plantation so this could change. No flooding impacts were recorded. Therefore, this woodland is still currently meeting its objectives.

3.3.5 Case study C 

3.3.5.1 Site description and background

Site C is a series of three plantations situated on a farm near the south bank of the River Dochart, Perthshire. Small streams run directly through the plantations from the main River Dochart (Figure 14). The woodland creation was established in 2020 through the FGS. The plantations are all deer-fenced, and trees are individually protected with vole guards (0.2 m tall). In total, the plantation measures 47.4 ha split across three main areas with around 75,000 trees planted in total. The woodland creation scheme is designed to enhance riparian woodland cover within an area where riparian woodland cover is sparse.

3.3.5.2 Beaver impacts

Beavers entered the plantations through two water gates (Figure 15) and impacts were evident. Evidence of gnawing through the water gates and fence posts was observed. A mix of fresh and old feeding signs were noted. NN methods estimated that 49% of the riparian zone across the three main areas was impacted by beaver browsing. A total of 98 trees had been felled by beaver, therefore at a plantation level, beaver browsing impacts totalled 0.1%. 

Notably, most of the browsed trees were cut just above the vole guards (Figure 16). Beavers had felled stems of willow (n=37), alder (n=25), rowan (n=21), birch (n=13), and aspen (n=2) at a mean diameter of 2.4 cm (±1.3 cm SD) reflecting the smaller trees present on this site and at an average height of 29.8 cm (±11.6 cm SD). There was no evidence of beaver flooding impacts.

3.3.5.3 Deer impacts

There were no impacts of browsing deer on the plantations and overall woodland herbivore browsing was recorded as ‘no impact.’ 

3.3.5.4 Overall factors influencing woodland objectives

Beaver browsing impacts were significant in the riparian zone of these plantations. Many planted stems were browsed at a young age and have not displayed signs of regeneration. However, it is estimated that this woodland is currently still reaching its targeted stocking densities overall and no browsing was recorded on young shoots of natural regeneration that was evident on site. Further study would be required as to whether the woodland that develops will provide all of the desired objectives, such as effects on water temperature, if tree and shrub cover in the riparian zone continues to be limited. 

4. Discussion

4.1 Case studies: lessons learned

One component of this study was to investigate the nature and scale of beaver impacts on existing woodland creation sites in Scotland. Beaver impacts were only recorded in three of the 14 woodland creation sites which is a small sample size to reliably characterise the nature and extent of beaver impacts. Furthermore, it was evident that each site presented its own unique set of variables, highlighting the context-dependent nature of beaver activity. Therefore, we cannot generalise too much from the limited number of cases, but we can make some observations. 

Tree protection, largely in the form of fencing, was helping to deter beaver and deer impacts. Our findings suggest that having an entry point (within a beaver territory) directly leading to a watercourse is likely to increase the likelihood of beaver impacts, most commonly via a water gate. If the goal is complete beaver exclusion, then it may be possible to adapt the location of the fence line to avoid creating a point of weakness. Alternatively, water gate specifications could be specially adapted to exclude beavers (see Appendix A for design and considerations) (Campbell-Palmer et al., 2016). However, the practicality of such measures and the additional costs are likely to limit the use of beaver-proof enclosures. 

Traditional tree guards appeared to be relatively ineffective at protecting planted trees from beaver impacts. Only one site had beaver and deer browsing so we cannot infer anything about deer management measures. 

Case Study A highlighted the importance of species selection as most of the trees had perished due to beaver-induced flooding. The site largely comprised oak trees, known to have low tolerance to flooding (Glenz et al., 2006). The consideration of suitable tree species extends not only to the riparian zone but any parts of a scheme which may be liable to damming and flooding due to the topography and existence of drainage channels that could be dammed or natural wetlands. The Ecological Site Classification (ESC) Tool (Forest Research, 2024b) already exists to help forest managers and planners select tree species that are ecologically suited to particular sites based on soil type. The potential for beavers to modify the water table on sites is a further consideration for the design of schemes. Wetland areas could either provide biodiversity benefits and/or require modifications to scheme design to ensure they are resilient to beaver impacts.

Case Study C comprised flood-tolerant species (birch, rowan, aspen, alder and willow) in the riparian zone which were subjected to high levels of beaver browsing impacts. Despite this, the overall impact on the plantation was arguably low. If the scheme in Case Study C had been entirely restricted to the riparian zone, then tree losses would have been proportionately greater. It remains to be seen if all the objectives for woodland creation e.g. casting shade on the river, will be delivered if this level of browsing persists. It seems likely that impacts on this site were heavily influenced by the sparseness of riparian woodland in the landscape. The literature from elsewhere in Europe indicates resource depletion is more likely where habitats are fragmented and sub-optimal and it may be more challenging to establish new woodlands in such situations. Natural regeneration may present a lower cost option to planting in such cases but would require similar considerations to ensure woodland creation objectives were met.

Despite our limited sample size, the beaver foraging preferences are consistent with other studies on (Jackowiak et al., 2020; Mikulka et al., 2022b), with alder having the highest number of browsed stems but, willow and aspen being most preferred relative to their availability. Although the focus of this study was on planted schemes, no beaver browsing impacts were recorded on natural regeneration within the woodland creation schemes. The average selected tree diameter (7 cm) and height of cut (35 cm) are similar to those reported in other research (with much larger datasets) (Iason et al., 2014; Janiszewski et al., 2017). 

As central-place foragers, beaver browsing is concentrated, clustered, and sporadic (Raffel et al., 2009; Piton et al. 2020). The nearest neighbour (NN) methods that are typically used for woodland damage assessments are based on measuring impacts on trees in equally spaced clusters, such that localised patches of high-intensity beaver browsing could be completely missed. Hence whilst NN methods may be effective for deer or squirrels (Gill, 1992), the sampling methods need to be modified for assessing beaver impacts. Walkovers surveys could be incorporated into existing monitoring of deer impacts before the start of the growing season (February to May) (NatureScot, 2022) and would also capture the effects of beaver browsing on woody material over the winter. 

As central-place foragers, beaver browsing is concentrated, clustered, and sporadic (Raffel et al., 2009; Piton et al. 2020). The nearest neighbour (NN) methods that are typically used for woodland damage assessments are based on measuring impacts on trees in equally spaced clusters, such that localised patches of high-intensity beaver browsing could be completely missed. Hence whilst NN methods may be effective for deer or squirrels (Gill, 1992), the sampling methods need to be modified for assessing beaver impacts. Walkovers surveys could be incorporated into existing monitoring of deer impacts before the start of the growing season (February to May) (NatureScot, 2022) and would also capture the effects of beaver browsing on woody material over the winter. 

4.2 Wider implications

Providing more extensive areas of high-quality, native riparian woodlands should result in less notable beaver impacts, a scenario akin to the large forests of central Europe (Parker & Rosell, 2003). The challenge in Scotland is to create this woodland resource whilst there are high levels of deer browsing and potentially locally high beaver impacts on smaller, more isolated woodland creation schemes. Questionnaire responses from European woodland managers suggest that beaver-deer interactions are considered a common occurrence, with the impact likely to depend on deer density (Wilson, 2022).

Some aspects of beaver browsing and damming can be predicted from foraging preference studies (Haarberg & Rosell, 2006) or beaver dam capacity models (Graham et al., 2020), however, specific site impacts cannot readily be predicted. Proactive consideration of potential beaver activity during the early design stages should allow for better outcomes. Providing woodland managers with information and planning tools would facilitate this process, including mapping of soils, topography and beaver dam capacity.

Beaver presence will require that some changes are made to forestry grants, standards and guidance. Inspections already allow some tolerance of deer impacts and a similar approach could be taken with regard to beaver impacts. Woodland managers may need to be prepared to ‘beat up’ where targets are not met and adapt plans in the event of sustained losses. Support for the additional costs of protecting trees from beaver impacts using fencing or individual tree protection could also be considered where this is necessary in higher risk locations. However, the wider goal should be to develop woodlands that are resilient to beaver herbivory, rather than resistant to it.

5. Recommendations

Given the anticipated restoration of beaver in Scotland both assisted by translocations and by natural dispersal, consideration needs to be given to how beavers will impact on woodland creation (and vice versa) at an early stage. Ultimately, not all beaver activity will be detrimental to woodland creation and may even enhance certain objectives. The current or future presence of beaver should not discourage the promotion of woodland creation even in more upland situations where riparian woodland is more fragmented and beaver impacts may be proportionately greater, because often these are the locations where woodland will best deliver multiple benefits. However, the successful design and creation of new woodland in the presence of beavers may require a different set of considerations than has traditionally been the case, principally with regard to managing deer impacts. However, effective deer management is likely to be key to establishing new woodlands where beavers are present. Figure 17 illustrates what a beaver resilient woodland may look like in comparison with Figure 1.

Based on the information presented in the literature review, insights gained from the questionnaire and observations from the site visits, the following recommendations are made.

• Setting woodland objectives: objectives need to be carefully considered for areas of a woodland creation scheme that beavers are likely to use; whether as forage in the riparian zone or that may be affected by beaver induced flooding. Non-productive objectives such as improving biodiversity, flood mitigation, carbon sequestration, increasing shade for fish are more likely to be compatible with beaver presence than commercial production in these locations. Consideration needs to be given at the outset to the level of beaver integration (none, partial, full). The approach taken will inform the overall design and approach to managing herbivore impacts.
• Monitoring and managing herbivore impacts: Based on the literature reviewed (and noting that the distances quoted vary slightly in individual studies), beaver impacts are typically concentrated within 20 m of the watercourse. Impacts may be proportionately greater in smaller schemes where there is less alternative riparian woodland in the landscape. Excluding beavers by fencing is likely to be practically challenging and costly. The appropriateness of deer and beaver fencing needs to be considered independently e.g. deer fencing could afford beaver browsed shoots the opportunity to regrow, allowing higher rates of establishment. However, there may be some situations where beaver fencing of small stands or individual tree protection is necessary to allow successful establishment. The practicalities of beaver inclusion or protection measures need to be further examined and considered in grant schemes, including specifications and conditions. Monitoring of beaver impacts should be incorporated into regular inspections for other herbivore impacts (deer, squirrel, hare etc). The extent of beaver-deer interactions should also be monitored and managed. Deer impacts will need to be managed at a level that will allow beaver impacts to be sustainable.
• Woodland design, species choice and stocking density: To increase scheme resilience, it is recommended that within this focal 20m of the watercourse, a diverse mix of beaver-preferred broadleaves species including willow, aspen, birch, rowan and smaller proportions of alder and naturalised bankside vegetation should be favoured. A riparian zone of fast-growing species such as willows and aspen should help to fulfil beaver foraging requirements and minimise impacts elsewhere within the scheme. Where a riverside strip of fast-growing species is planted, a stocking density of 1000 stems per ha or greater is advised. Again, based on findings in the literature, trees planted beyond 50 m of the watercourse are unlikely to be impacted by beaver foraging and can be matched to site type and woodland objectives; taking account of the potential for waterlogging in low-lying areas. Where possible, the scale of the scheme should take account of beaver foraging requirements, with the establishment of a strip or patches of favoured tree and shrub species replicated along the watercourse (suggested at 0.1 ha per km of watercourse), allowing for spatial and temporal cycling of beaver cutting and regrowth.
• Adaptive management: woodland managers should be prepared to have to respond to beaver impacts, whether through enhanced deer management or mitigation of flooding impacts, however, proactive planning should help to reduce conflicts.

5.1 Further development/ research needs

Whilst the aim of this study was to inform guidance for woodland creation in the presence of beavers, it is clear there are some areas that required further research before evidence-based recommendations can be made. The following key areas for further development are proposed.

• The literature review highlighted the potential benefit of promoting fast growing willow and or aspen and naturalised bankside vegetation along the watercourse to fulfil beaver foraging requirements and reduce impacts on the adjacent woodland. Hence it would be helpful to confirm recommendations for the scale and distribution of riverside woodland creation required to meet these requirements (0.1 ha or (10 m x 10 m) per km of watercourse from Mikulka et al. 2022a). Further trials would help to inform the practicalities of these recommendations at woodland creation sites in Scotland, particularly where there is little existing woodland in the landscape, including consideration of methods of establishment via natural regeneration or planting e.g. of willow whips.
• Trialling of individual tree protection designs that are effective at deterring beaver impacts as an alternative to plastic tree guards. For example, a guard that incorporates the gritty texture of deterrent paint and the microclimate of a plastic tree guard may be worth trialling. The utility and cost effectiveness of such an approach would also need to be further explored.
• Woodland planners would benefit from the development of decision support tools that help to assess the potential for future beaver impacts through changes to hydrology. Models could help to identify locations within a woodland creation site that would better deliver environmental objectives, e.g. wet woodland or standing water. Such models might usefully incorporate information from ESC, topographic data and beaver dam capacity modelling.
• Decision makers could benefit from modelling which assesses and maps the risk of specific proposals for woodland creation and for strategic planning. A risk-based analysis could be based on information on beaver territories and the extent and distribution of woodland habitats in the landscape. High risk sites may still be strong candidates for woodland creation to deliver multiple benefits but may require additional consideration of how this is achieved.
• As already highlighted, woodland herbivore impact assessment methods and damage assessment methods need to be updated to incorporate beaver impacts and beaver-deer interactions to enable appropriate management responses to be taken. Any changes to existing methods would require field testing.

6. References

Armstrong, G.S., Aprahamian, M.W., Fewings, G.A., Gough, P.J., Reader, N.A. & Varallo, P.V. 2010. Environment agency fish pass manual: guidance notes on the legislation, selection and approval of fish passes in England and Wales. Environment Agency, Rio House, Bristol, p.369.

Armstrong, H., Black, B., Holl, K., & Thompson, R. 2023. The Woodland Herbivore Impact Assessment Method User Guide.

Arnold, J.M., Gerhardt, P., Steyaert, S. M. J. G., Hochbichler, E., & Hackländer, K. 2018. Diversionary feeding can reduce red deer habitat selection pressure on vulnerable forest stands but is not a panacea for red deer damage. Forest Ecology and Management, 407, 166–173.

Bateman, H.L., & Merritt, D.M. 2020. Complex riparian habitats predict reptile and amphibian diversity. Global Ecology and Conservation, 22, p.e00957.

Campbell-Palmer, R., Gow, D., Schwab, G., Halley, D.J., Gurnell, J., Girling, S., Lisle, S., Campbell, R., Dickinson, H., & Jones, S. 2016. The Eurasian beaver handbook: ecology and management of Castor fiber. Pelagic Publishing Ltd.

Campbell-Palmer, R., Puttock, A., Wilson, K.A., Leow-Dyke, A., Graham, H.A., Gaywood, M.J., & Brazier, R.E. 2021. Using field sign surveys to estimate spatial distribution and territory dynamics following reintroduction of the Eurasian beaver to British river catchments. River Research and Applications, 37(3), 343–357.

Carpio, A.J., Apollonio, M., & Acevedo, P. 2021. Wild ungulate overabundance in Europe: contexts, causes, monitoring and management recommendations. Mammal Review, 51(1), 95–108.

Cole, L.J., Stockan, J., & Helliwell, R. 2020. Managing riparian buffer strips to optimise ecosystem services: A review. Agriculture, Ecosystems and Environment, 296 (April), 106891.

Corbacho, C., Sánchez, J.M., & Costillo, E. 2003. Patterns of structural complexity and human disturbance of riparian vegetation in agricultural landscapes of a Mediterranean area. Agriculture, Ecosystems & Environment, 95(2–3), 495–507.

Crawley, M.J. 1984. Herbivory: the dynamics of animal-plant interactions. Studies in Ecology. In Progress in Physical Geography: Earth and Environment (Vol. 10). Blackwell Scientific Publications.

DECS (Deer Commission Scotland, Forestry Commission Scotland, Scottish Natural Heritage & Scottish Executive Environment and Rural Affairs Department), 2004. Joint Agency Statement and Guidance on Deer Fencing.

Eason, W.R., Gill, E.K. & Roberts, J.E. 1996. Evaluation of anti-sheep tree-stem-protection products in silvopastoral agroforestry. Agroforestry systems, 34, 259-264.

Forest Research, 2024a. Creating and managing riparian woodland. UK Forestry Practice Guide. 

Forest Research, 2024b. Tools and Resources: Ecological Site Classification (ESC). 

Forestry Commission, 2021. A guide to planning new woodland in England Woodland Creation Planting for the future. 

FOWARA, 2006. Forested Water Retention Areas: Guidelines for decision makers, forest managers and landowners. Edited by: Armbruster, J., Muley-Fritze, A., Pfarr, U., Rhodius, R., Siepmann-Schinker, D., Sittler, B., Späth, V., Trémolières, M., Rennenberg, H. & Kreuzwieser, J. Institute for Landscape Ecology and Nature Conservation, Buehl, Germany.

Fuentes-Montemayor, E., Goulson, D., Cavin, L., Wallace, J.M., & Park, K.J. 2013. Fragmented woodlands in agricultural landscapes: The influence of woodland character and landscape context on bats and their insect prey. Agriculture, Ecosystems & Environment, 172, 6–15.

Fuentes‐Montemayor, E., Ferryman, M., Watts, K, Macgregor, N.A., Hambly, N., Brennan, S., Coxon, R., Langridge, H., & Park, K.J. 2020. Small mammal responses to long‐term large‐scale woodland creation: the influence of local and landscape‐level attributes. Ecological Applications, 30(2), e02028.

Gaywood, M.J., Stringer, A., Blake, D., Hall, J., Tree, A., Genney, D., Macdonald, I., Tonhasca, A., Bean, C., Mckinnell, J., Raynor, R., Watkinson, P., Bale, D., Scott, J., & Blyth, S. 2015. Beavers in Scotland: A report to the Scottish Government. Scottish Natural Heritage, Inverness.

Gerwing, T.G., Johnson, C.J., & Alström-Rapaport, C. 2013. Factors influencing forage selection by the North American beaver (Castor canadensis). Mammalian Biology, 78(2), 79–86.

Gill, R.M.A. 1992. A review of damage by mammals in north temperate forests: 3. Impact on trees and forests. Forestry, 65(4), 363–388.

Gill, R.M.A., & Morgan, G. 2010. The effects of varying deer density on natural regeneration in woodlands in lowland Britain. Forestry, 83(1), 53-63.

Glenz, C., Schlaepfer, R., Iorgulescu, I., & Kienast, F. 2006. Flooding tolerance of Central European tree and shrub species. Forest Ecology and Management, 235(1–3), 1–13.

Graham, H.A., Puttock, A., Macfarlane, W.W., Wheaton, J.M., Gilbert, J.T., Campbell-Palmer, R., Elliott, M., Gaywood, M.J., Anderson, K., & Brazier, R.E. 2020. Modelling Eurasian beaver foraging habitat and dam suitability, for predicting the location and number of dams throughout catchments in Great Britain. European Journal of Wildlife Research, 66(3).

Haarberg, O., & Rosell, F. 2006. Selective foraging on woody plant species by the Eurasian beaver (Castor fiber) in Telemark, Norway. Journal of Zoology, 270(2), 201–208.

Hall, J., Pollock, M.M. & Hoh, S. 2011. Methods for successful establishment of Cottonwood and Willow along an incised stream in semi-arid Eastern Oregon, USA. Ecological Restoration, 29(3), 261-269.

Hall, J.E., Kirby, K.J. & Whitbread, A.M. 2004. National vegetation classification field guide to woodland, Joint Nature Conservation Committee (JNCC), Peterborough.

Hall, J.E., Pollock, M.M., Hoh, S., Volk, C., Goldsmith, J., & Jordan, C.E. 2015. Evaluation of dryland riparian restoration with cottonwood and willow using deep‐planting and herbivore protection. Ecosphere, 6(12), 1-12. 

Harmer, R., Watts, K., & Ray, D. 2015. A hundred years of woodland restoration in Great Britain: changes in the drivers that influenced the increase in woodland cover. Restoration of boreal and temperate forests. Second edition. CRC Lewis, Boca Raton, Florida, USA, 299-320.

Hodge, S., & Pepper, H. 1998. The prevention of mammal damage to trees in woodland. Forestry Practice Note. Forestry Commission. 

Hood, G.A., & Bayley, S.E. 2008.The effects of high ungulate densities on foraging choices by beaver (Castor canadensis) in the mixed-wood boreal forest. Canadian Journal of Zoology, 86(6), 484–496.

Hyvönen, T., & Nummi, P. 2008. Habitat dynamics of beaver Castor canadensis at two spatial scales. Wildlife Biology, 14(3), 302–308.

Iason, G.R., Sim, D., Brewer, M.D., & Moore, B.D. 2014. The Scottish Beaver Trial: woodland monitoring 2009-2013, final report. Scottish Natural Heritage Commissioned Report No. 788.

Ivlev, V.S. 1962. Experimental Ecology of the Feeding of Fishes. Copeia, 1962(1), 234.

Jackowiak, M., Busher, P., & Krauze-Gryz, D. 2020. Eurasian beaver (Castor fiber) winter foraging preferences in northern Poland—the role of woody vegetation composition and anthropopression level. Animals, 10(8), 1–12.

Janiszewski, P., Kolasa, S., & Strychalski, J. 2017. The preferences of the European beaver Castor fiber for trees and shrubs in riparian zones.

Johnston, C.A. 2017. Beaver Loggers: Beaver Herbivory Alters Forest Structure. Beavers: Boreal Ecosystem Engineers, 91–116.

Jones, K.C., Gilvear, D., Willby, N.J., & Gaywood, M. 2009. Willow (Salix spp.) and aspen (Populus tremula) regrowth after felling by the Eurasian beaver: implications for riparian woodland conservation in Scotland. Aquatic Conservation: Marine and Freshwater Ecosystems, 19, 75–87.

Justice, C., White, S.M., McCullough, D.A., Graves, D., & Blanchard, M.R. 2017. Can stream and riparian restoration offset climate change impacts to salmon populations? Journal of Environmental Management, 188, 212–227.

Kamczyc, J., Bielachowicz, M., & Pers-Kamczyc, E. 2016. Damages caused by European beaver (Castor fiber L., 1758) in broadleaved stands. Forestry Letters, 109, 7–10.

King County, 2022. Planning for beavers manual: anticipating beavers when designing restoration projects. Prepared by Jen Vanderhoof, King County Water and Land Resources Division. Seattle, Washington.

Kivinen, S., Nummi, P., & Kumpula, T. 2020. Beaver-induced spatiotemporal patch dynamics affect landscape-level environmental heterogeneity. Environmental Research Letters, 15(9). 

Lenane, R. 2016. Keeping Rivers Cool: A Guidance Manual. Creating riparian shade for climate change adaption. Woodland Trust. 

Loeb, R.E., & Garner, T.B. 2022. Natural Forest Regeneration Changes in an Urban Natural Area Forest with White-Tailed Deer (Odocoileus virginianus) Exclusion and Felling by North American Beaver (Castor canadensis). Natural Areas Journal, 42(3), 252–256.

Loeb, R.E., King, S., & Helton, J. 2014. Human pathways are barriers to beavers damaging trees and saplings in urban forests. Urban Forestry and Urban Greening, 13(2), 290–294.

Martell, K.A., Foote, A.L., & Cumming, S.G. 2006. Riparian disturbance due to beavers (Castor canadensis) in Alberta’s boreal mixedwood forests: Implications for forest management. Ecoscience, 13(2), 164–171.

Miistakis Institute, 2021. Textural beaver repellent for tree protection.

Mikulka, O., Adamec, Z., Kamler, J., Homolka, M., Drimaj, J., Plhal, R., & Petr, P. 2022a. Using deciduous softwoods to protect commercial forest stands against damage by Eurasian beaver (Castor fiber L.). Forest Ecology and Management, 520.

Mikulka, O., Homolks, M., Drimaj, J., & Kramler, J. 2022c. Feeding behaviour of Eurasian beavers (castor fiber) along small streams in an agricultural landscape. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, 70(2), 71-82

Mikulka, O., Pyszko, P., Skoták, V., Kamler, J., Drimaj, J., Plhal, R. and Homolka, M. 2022b. The Influence of Forestry Management on the Selection of a Non-Vegetative Diet by the Eurasian Beaver (Castor fiber L.). Animals, 12(21), 2949.

Misiukiewicz, W., Gruszczyńska, J., Grzegrzółka, B., & Januszewicz, M. 2016. Impact of the European beaver (Castor fiber L.) population on the woody vegetation of Wigry National Park. Roczniki Naukowe Polskiego Towarzystwa Zootechnicznego, 12(3), 45–64.

NatureScot, 2020. Protecting trees from beavers using wire mesh: beaver practical mitigation guides. 

NatureScot, 2022. Wild Deer Best Practice Guidance for Scotland.

NatureScot, 2023. Riparian woodland by Local Authority Map (percentage of the search area that is woodland).

Nolet, B.A., Hoekstra, A., & Ottenheim, M.M. 1994. Selective foraging on woody species by the beaver Castor fiber, and its impact on a riparian willow forest. Biological conservation, 70(2), 117-128.

Nolet, B. A. 1997. Management of the beaver (Castor fiber): toward restoration of its former distribution and ecological function in Europe. Council of Europe Nature and Environment Report No. 86. Council of Europe Publishing, Strasbourg.

Nolte, D.L., Lutman, M.W., Bergman, D.L., Arjo, W.M., & Perry, K.R. 2003. Feasibility of non-lethal approaches to protect riparian plants from foraging beavers in North America. USDA National Wildlife Research Center-Staff Publications, 256. 

Nummi, P., & Kuuluvainen, T. 2013. Forest disturbance by an ecosystem engineer: Beaver in boreal forest landscapes. Boreal Environment Research.

Ogilvy, T., Melville, N., Martinez Contributing, R., Martinez, R., Stutter, M., Sime, I., Wilkinson, M., Morison, J., Broadmeadow, S., & O’Brien, L. 2022. Riverwoods for Scotland - Report on Scientific Evidence. 

Parker, H., & Rosell, F. 2003. Beaver management in Norway: a model for continental Europe? Lutra, 46(2), 223–234.

Parrott, J. & MacKenzie, N. 2009. Aspen in Scotland: biodiversity and management. Proceedings of a Conference held in Boat of Garten, Scotland 3-4 October 2008. 

Pejstrup, M.S., Andersen, J.R., & Mayer, M. 2023. Beaver foraging patterns in a human-dominated landscape: Effects on woody vegetation and mammals. Forest Ecology and Management, 528, 120645.

Pepper, H. 1998. Nearest Neighbour Method for Quantifying Wildlife Damage to Trees in Woodland.

Pepper, S., Barbour, A., & Glass, J. 2020. The Management of Wild Deer in Scotland: Report of the Deer Working Group. Scottish Government, Environment and Forestry Directorate. 

Piton, G., Loucougaray, G., Daumerguew, N., Breton, V., & Evette, A. 2020. The beaver’s menu: species and spatial selection of a European beaver population and implications for riverbank bioengineering. Wetlands Ecology Management, 28(4), 901-908.

Pollock, M., Lewallen, G., Woodruff, K., Jordan, C., & Castro, J. 2015. The Beaver Restoration Guidebook: A Practitioners Guide to Working with Beaver to Restore Streams, Wetlands and Floodplains. United States Fish and Wildlife Service, 219p.

Puttock, A., Graham, H.A., Ashe, J., Luscombe, D. J., & Brazier, R.E. 2021. Beaver dams attenuate flow: A multi‐site study. Hydrological processes, 35(2), e14017.

Puttock, A., Graham, H.A., Cunliffe, A.M., Elliott, M., & Brazier, R.E. 2017. Eurasian beaver activity increases water storage, attenuates flow and mitigates diffuse pollution from intensively-managed grasslands. Science of the total environment, 576, 430-443.

Raffel, T.R., Smith, N., Cortright, C., & Gatz, A.J. 2009. Central Place Foraging by Beavers (Castor canadensis) in a Complex Lake Habitat. 162(1), 62–73.

Ramirez, J.I., Jansen, P.A., & Poorter, L. 2018. Effects of wild ungulates on the regeneration, structure and functioning of temperate forests: A semi-quantitative review. Forest Ecology and Management, 424(May), 406–419.

Redick, C.H., & Jacobs, D.F. 2020. Mitigation of deer herbivory in temperate hardwood forest regeneration: A meta-analysis of research literature. Forests, 11(11), 1–16.

Reeves, J.T., Ezell, A.W., Hodges, J.D., Schultz, E.B., & Self, A.B. 2016. Influence Of Flooding, Freezing, And American Beaver Herbivory on Survival of Planted Oak Seedlings. Silvicultural Research Conference, 465.

Reimoser, F., & Putman, R. 2011. Impacts of wild ungulates on vegetation: costs and benefits. Ungulate management in Europe: problems and practices. Cambridge University press, Cambridge, 144-191.

Reynolds, P. 2000. European beaver and woodland habitats. Scottish Natural Heritage Review No. 126.

Rood, S.B., Bigelow, S.G., Polzin, M.L., Gill, K.M., & Coburn, C.A. 2015. Biological bank protection: trees are more effective than grasses at resisting erosion from major river floods. Ecohydrology, 8(5), 772–779.

Rosell, F., Bozser, O., Collen, P., & Parker, H. 2005. Ecological impact of beavers Castor fiber and Castor canadensis and their ability to modify ecosystems. Mammal Review, 35(3–4), 248–276.

Rosell, F., & Campbell-Palmer, R. 2022. Beavers: ecology, behaviour, conservation, and management. Oxford University Press.

Scottish Forestry, 2020. Preparing Woodland Creation Applications - A guide for land managers. 

Sing, L., Towers, W., & Ellis, J. 2013. Woodland expansion in Scotland: an assessment of the opportunities and constraints using GIS. Scottish Forestry, 67(4), 18-25.

Spake, R., Bellamy, C., Gill, R.M.A., Watts, K., Wilson, T., Ditchburn, B., & Eigenbrod, F. 2020. Forest damage by deer depends on cross-scale interactions between climate, deer density and landscape structure. Journal of Applied Ecology, 57(7), 1376–1390.

Spitzer, R., Felton, A., Landman, M., Singh, N.J., Widemo, F., & Cromsigt, J.P.G.M. 2020. Fifty years of European ungulate dietary studies: a synthesis. Oikos, 129(11), 1668–1680.

Staton, T., Beauchamp, K., Broome, A., & Breeze, T. 2024. Tree Species Guide for UK Agroforestry Systems. University of Reading & Forest Research.

Stringer, A.P., & Gaywood, M.J. 2016. The impacts of beavers Castor spp. on biodiversity and the ecological basis for their reintroduction to Scotland, UK. Mammal Review, 46(4), 270-283.

Tanentzap, A.J., Zou, J., & Coomes, D.A. 2013. Getting the biggest birch for the bang: Restoring and expanding upland birchwoods in the Scottish Highlands by managing red deer. Ecology and Evolution, 3(7), 1890–1901.

Thomas H., Paterson J., Metzger M., Sing L. 2015. Towards a research agenda for woodland expansion in Scotland. Forest Ecology and Management, 349, 149–161.

Trout, R.C. & Pepper, H.W. 2006. Forest fencing. Forestry Commission Technical Guide. Forestry Commission, Edinburgh. 

Turunen, J., Markkula, J., Rajakallio, M., & Aroviita, J. 2019. Riparian forests mitigate harmful ecological effects of agricultural diffuse pollution in medium-sized streams. Science of The Total Environment, 649, 495–503.

Turunen, J., Elbrecht, V., Steinke, D., & Aroviita, J. 2021. Riparian forests can mitigate warming and ecological degradation of agricultural headwater streams. Freshwater Biology, 66(4), 785–798.

Valachovic, D. 2014. Manual of beaver management within the Danube river basin. Danube Parks, 91.

Vorel, A., Válková, L., Hamšíková, L., Maloň, J., & Korbelová, J. 2015. Beaver foraging behaviour: Seasonal foraging specialization by a choosy generalist herbivore. Behavioural Ecology and Sociobiology, 69(7), 1221–1235.

Vorel, A.A., Dostál, T., Uhlíková, J., Korbelová, J., & Koudelka, P. 2016. Handbook for Coexisting with Beavers. Czech University of Life Sciences, Prague, 1-137.

Wazna, A., Cichocki, J., Bojarski, J., & Gabryś, G. 2018. Selective foraging on tree and shrub species by the European beaver Castor fiber in lowland and highland habitats in Western Poland. Polish Journal of Ecology, 66(3), 286–300.

Wróbel, M. 2020. Population of Eurasian beaver (Castor fiber) in Europe. Global Ecology and Conservation, 23, e01046.

Westbrook, C.J. 2021. Beaver as agents of plant disturbance. In: Plant disturbance ecology. Academic Press. pp. 489-528.

Westbrook, C.J., & England, K. 2022. Relative Effectiveness of Four Different Guards In Preventing Beaver Cutting of Urban Trees. Environmental Management, 70(1), 97–104.

Wilson, K.A. 2022. Interactive Effects of Beaver and Deer on Scotland’s Riparian Woodlands. PhD Thesis. University of Stirling, Stirling. 

Wilson, K.A., Law, A., Hester, A.J., Gaywood, M.J. & Willby, N. 2024. Tree felling by beaver promotes regeneration in riparian woodlands whilst increasing resource availability for deer. Forest Ecology and Management, 562, 121910. 

Woodland Trust, 2022. Tree species handbook: A technical guide for practitioners. 

Ziegltrum, G.J. 2004. Efficacy of black bear supplemental feeding to reduce conifer damage in western Washington. Journal of Wildlife Management, 68, 470–474.

Appendices

Appendix A. Beaver exclusion specifications

A1. Beaver exclusion specifications - fencing

The following specifications have been trialled and tested throughout >30 successful beaver enclosures throughout Britain. Although originally designed for beaver enclosures, they should be adapted as necessary to suit woodland creation schemes. Specifications for the fence type recommended have been devised after consultation with Tornado wire, however there are other suppliers of different fence mesh types. Critical to their use are the basic dimensions and specification outlined below. 

The fencing should generally be Tornado RL19/180/5 C14 mesh which should extend 120 cm vertically up the timber work with a 50 cm section at the top set at 45 degrees outside of the plantation as an anti-climb device for a minimum of 20 m on either side of an in or outflow and for any areas likely to flood as outlined above. This fence style should have a sunken element extending down into the ground for 122 cm or as far as bedrock allows. The materials for this section should comprise 122 cm by 244 cm weldmesh panels with 5 cm square apertures. At the base of the fence an anti-dig curtain of Tornado RL6/50/5 C5 mesh should be laid on the surface of the ground attached to the base of the vertical mesh with hog rings and ground pegged to the surface where it should extend 90 cm into the enclosure on the floor. The wire mesh sizes allow for an overlap of at least 10 cm preventing beaver access at the join.

Around the perimeter of the beaver enclosure where the fence is well away from any water course or wet ground and it is very unlikely any beavers should attempt to climb. Tornado R13/120/5 mesh should extend 120 cm vertically up the timber work, with Tornado RL6/50/5 C5 mesh used an anti-dig precaution laid on the floor extending 90 cm into the enclosure and overlapping the vertical mesh by 10 cm. This should be fixed to the floor with pegs and hog ringed to the vertical wire as above.

The timber should be placed on the outside of the fence so that it is protected by the steel mesh. It should overlap the vertical mesh by 10 cm. This should be fixed to the ground with pegs and hog ringed to the vertical wire as above. Any trees likely to fall onto the fence should either be wrapped in wire meshed to prevent beaver feeding or felled.

A2. Beaver exclusion specifications – water gates

The following specifications have been devised to protect waterways in beaver enclosures throughout Britain. They may be adapted as necessary for woodland creation sites. 

A rectangular frame made of 2-inch square steel should be bespoke made for the dimensions of the area and placed into the stream. This feature should have 10-millimetre (mm) spacings along its length top and bottom which allow for a metal reinforcing bar to be inserted through and then tapped down into the stream bed. This process should enable the bar to be inserted into the bed for a depth of 100 cm or until it hits bedrock. The bar should then be secured at the top with a bolt. This system involves minimal ground disturbance and should be checked twice weekly or after any storm or flood event for clearance. The top sections of the grill should be interconnected to the fence. Where the grill sits in the stream the sides and base of the channel should be covered with a mesh screen and weld-mesh sunk to a depth of 1.2m - as practicable according to the bedrock level - should be installed to 20 m along the fence line on either side of the grill.

Such water gate designs have been used in >30 beaver enclosures in Britain in order to retain beavers. Larger gaps (i.e., more than 10 cm x 10 cm) have enabled access to younger individuals and beaver fatalities have been reported in Europe and North America from animals getting trapped in fencing. Before such structures are installed, consideration for migratory fish must be given as the current guidance suggests that grill spacing should be no less than 20 cm when Atlantic salmon (Salmo salar) are present and no less than 15 cm for sea trout (Salmo trutta) (Armstrong et al., 2010). Given this guidance, it may not be feasible to install water gates on watercourses with migratory fish. 

Appendix B. Tree protection guidance

Appendix C. Use of beaver deterrent paint 

Application of the ‘Wöbra®’ commercial anti-game paint to bark on the lower tree trunk has shown documented success in deterring browsing sheep (Eason et al., 1996), but there is currently no peer-reviewed literature that demonstrates its use against beaver browsing. However, the manufacturers have provided NatureScot with information on Dutch and Austrian studies of efficacy (Bryce pers comm²). One of these studies included trials across four sites in two regions of Germany of the most commercially important tree species in the area (600 beech trees, 452 oak, 170 birch and 289 ‘noble deciduous trees’, (i.e., not the main commercially grown hardwoods such as maples, elms, ash, cherry, rowan, fruit trees). They report that when browsing pressure is high, WOBRA can reduce damage by 30% compared to unprotected trees and that full protection is dependent on their being enough alternative food available. WOBRA is currently not a licenced product in the UK, though NatureScot have permission to trail its use under a permit from the Health and Safety Executive. 

Current costs of WOBRA (including import) are equivalent or more expensive than using mesh guards and is slightly more time consuming to apply (NatureScot per com.). Application is currently limited to selective use on high value trees that are not suitable for protection using mesh due to the tree form or other aesthetic considerations. However, as the application first requires removal of any vegetation growing on the bark, the use of deterrent paint is not suitable for trees that are host notable lichen or bryophyte communities. This tends to limit its use to smooth barked species lacking crevices that may host such species.

A short-term pilot study by the Miistakis Institute (2021) tested the effectiveness of ‘DIY’ abrasive paints (a mix of masonry paint and sand) and found them to be effective. However, there is some concern regarding the toxicity of latex paint to beavers (and the aquatic environment). Long-term effectiveness is likely to be dependent on repeat applications (height and thickness of application, coverage of exposed roots and bark roughness). Many North American websites highlight that deterrent paint is less effective on trees of small diameter/ saplings, it is also suggested that the continued growth of saplings that are less than 6 feet tall, will limit the repellent’s effectiveness and longevity. 

² The product was first approved by the Brauns company in 1986 as an anti-peeling agent against red deer after three years of testing and approval procedures. As early as 1993, there were trials on the use of Wöbra against beavers as part of a diploma thesis in forestry and in 1997 by the State Institute for Agriculture, with success, and a further diploma thesis in 1998. Approval against beavers was applied for in 1998 and granted in 2000. For several years now, Wöbra has been recommended as a long-term protection against beavers in Germany and Austria by beaver representatives (from nature conservation organizations and nature conservation authorities). Due to its environmental compatibility, the product has also been included in the FiBL list (list of inputs for organic farming) and has thus greatly extended the range of applications to fruit trees in particular. Our main customers for Wöbra are in Germany and Austria, but also in the Czech Republic, Switzerland, the Netherlands, France, Belgium and Slovenia’. 

Appendix D. Questionnaire information

D1. Questionnaire letter

The following background information was provided:

Scoping of advice in relation to beavers and riparian woodland creation - Survey Information & Questions

NatureScot (Scotland's nature agency) has commissioned a short study on beavers and riparian woodland creation. Overall, the study aims to provide insight into the impact of beavers and how land managers can design woodland creation schemes that take their presence into account. 

The study is being carried out by Dr Kelsey Wilson and Dr Róisín Campbell-Palmer through nature restoration organisation The Beaver Trust. There are three main parts:

• Investigating the impact of beavers on existing woodland creation sites in Scotland, assessing the scale and type of the impact whilst noting the presence of dams and associated flooding.
• Conducting a literature search of forestry practice in relation to woodland creation/restocking in countries where beavers are present, and how their presence affects such schemes. This includes assessing herbivore impacts and whether there is guidance available for managing these impacts.
• Using the information from the previous parts to make recommendations for the design of riparian woodland creation schemes, taking into account the objectives of the woodland creation and the influence of beaver activity on the hydrology of low-lying areas.

As part of the project’s objectives, we are combining data collected from case studies in Scotland with insight from forestry/land managers in Europe to assess the role of beavers in new woodland creations (or the restocking of existing woodlands) in the riparian zone.

Questionnaire closing date of 15th July 2023.

D2. Questionnaire questions

Participants were asked the following 20 questions:

1. Name
2. Organisation
3. Role
4. Are you familiar with the concept of incorporating beavers as a natural element in woodland creation?
5. Have you encountered any forestry practice guidance or recommendations in your country that addresses the use of beavers as a natural element in woodland creation?
6. What forestry practice guidance or recommendations have you encountered? Please include links or document titles.
7. In your opinion, are there benefits that beavers can bring to woodland creation and management?
8. In your opinion, what are the benefits beavers can bring to woodland creation and management?
9. In your opinion, are there negatives that beavers can bring to woodland creation and management?
10. In your opinion, what are the negatives beavers can bring to woodland creation and management?
11. Have you ever considered incorporating beavers into your woodland management planning?
12. Do you believe that there is sufficient information in current woodland management guidance to address the impacts of beavers on woodlands, particularly in relation to herbivore management?
13. Have you ever encountered situations where there were interactions between beavers and other herbivores (e.g., deer) in woodlands you manage?
14. If yes, how were these situations managed?
15. Are there any measures that you have implemented (or would recommend) to manage the impact of beavers on woodland creation and management?
16. Which aspect do you think has the bigger impact on woodland – beaver tree-felling or beaver-induced flooding?
17. What scale of impact do you think beaver tree-felling has on woodland?
18. What scale of impact do you think beaver induced flooding has on woodland?
19. Which aspect do you think has the bigger impact on woodland – beaver tree-felling or deer browsing?
20. Are there any additional comments or information you would like to share on beavers and riparian woodland creation?

D3. Participant responses to open questions (previously listed in Appendix D2)

Grammar and spelling was improved only as necessary to improve readability.

Benefits beavers can bring to woodland creation and management:

• Beavers coppice trees, helping to create healthy and diverse woodland.
• Dynamic changes in canopy structure and opening up the ground for wider changes in structure and ground flora to develop. This will have all the benefits associated with anthropomorphic woodland management in terms of a greater diversity of habitats and opportunities for a wide suite of wildlife.
• Differentiation of the structure, restoration of the suitability for woody species typical of the floodplain by rewetting, creation of open ground as a germination bed
• They create a very natural riparian forest along river valleys.
• These habitats created by beavers are very diverse, dynamic and very species rich and there are high population densities of different groups.
• The restructuring of whole forests to create dynamic riparian environments with increasing amounts of standing water which is otherwise absent as an ecotope from our national norm of forest ecology. Additional impacts resulting from this include enhanced biodiversity including the return of lost species, a much greater biomass of biodiversity and the sustainable reoccurrence of novel habitats such standing and sunken dead wood.
• Beavers in the long term reignite the process of natural succession of woodland to wetland to meadow to woodland and wetland again. This eternal cycle with all its different phases simply cannot be without beavers.  

Negatives beavers can bring to woodland creation and management:

• If deer numbers are too high their secondary browsing may prevent coppicing but this is more a land management issue than a beaver issue.
• Beavers could impact newly created woodland by browsing hard on trees before they’ve had time to establish.
• Strictly speaking beavers are a nuisance in productive forestry - mainly because of damming in low lying areas. The direct impact of felling trees is a minor problem.
• Whether 'negative' is the right word I don't know as it all comes down to perceptions. But if the ambition is to create a closed canopy commercial woodland then the impacts of inundation in low lying areas could be significant. Feeding and foraging might have more minor impacts but could still restrict smaller blocks of woodland from getting established. I think the latter will depend on availability of other food resources nearby. In Europe I have seen areas where flooding of woodland has led to the death of multiple stands of trees and I can imagine that if your ambition was a woodland for carbon storage or commercial use this would be perceived as 'negative'. Dead wood is obviously a brilliant habitat and lacking across many of our landscapes but it is still often perceived as 'bad' or 'negative' by members of the public or land managers we work with in Devon.
• They can flood areas of commercial conifers which are planted in drained valley systems, ring bark both confers and some valuable deciduous tree species such as beech and impact forest infrastructures such as roads or culverts.
• Some private landowners claim that the presence of beavers decrease the number of deer (both roe and red) due to flooding of feeding grounds. Dams are being reduced - but not if the "hunting" argument is the main case.
• I do not manage woodland. But I have observed where beaver coppiced willows have been foraged on by deer at beaver sites across Devon. I can imagine that, depending on deer pressure in an area, this could restrict the regeneration of the coppice regrowth as time goes on. Tree felling by beavers and then deer grazing of re-growth could therefore lead to undesired consequences in terms of understorey establishment and regeneration.
• No management necessary. Deer concentrate browsing at beaver sites, this reduces conflicts in other parts of the forest.
• Shooting the deer. 

Measures that respondents have implemented (or would recommend) to manage the impact of beavers on woodland creation and management:

• We are exploring the option of over planting willow for new planting, as well as looking at fencing off newly planted trees.
• If the aim of management of an area is nature restoration - the beavers fit perfectly. If the aim is forestry production - beavers are to be handled according to existing rules and regulations.
• "We are currently working with the Forestry Commission in England to try to incorporate details of managing potential beaver impacts into the English Woodland Creation Offer grant scheme. At the moment there is no clear guidance on how to maximise the benefits beavers could bring to riparian woodland management whilst managing concerns. We are working with them to hopefully develop this guidance and trial measures at pilot sites to explore how certain management could achieve desired outcomes. These measures are likely to include:
  1. Using modelling outputs from the University of Exeter to consider likely impacts and risks depending on proximity to other blocks of suitable habitat and potential capacity for beavers to build dams.
  2. Zonation of planting (targeted willow/poplar planting along water course edge and planting species which favour drier areas further back from waters edge).
  3. Protecting specific (using wire mesh) or blocks of trees (using fencing) depending on the objective of management. E.g. for amenity, commercial or specific carbon storage objectives.
• Creating large willow bushes along the shoreline as preferred and distracting food.
• The Swiss government has implemented the beavers into humid forest creation or "riparian forest" creation. We built a beaver floodplain model for the whole country to see where beavers can bring water on the floodplain and flood forests. In the national program for nature protection. The floodplain model is a base for the cantonal authorities to plan humid forest.
• Depends what you want to do if for example you use the example of the Altmuhl River in Bavaria the dense planting of willow to stabilise the banks of the re meandered river provided beavers with the opportunity in time to harvest this resource and complicate the water course still further. If you want to deter beaver impact in a given area of the riparian zone you could plant alder or other species like bird cherry within their foraging zone to maintain trees that are largely unpalatable.   

Additional comments or information you would like to share on beavers and riparian woodland creation:

• "Regarding the above questions on impact I think it is so context specific it will be hard to tell and therefore will require unique planning for each site. I suspect the scale of impact will depend on:
  1. The location of planting. E.g. the proximity to the watercourse and size of the water course will have an influence. Next to large watercourses (e.g. >5 order streams) we know it is unlikely beavers will build sustainable dams and inundate the adjacent land. Here we can be more confident inundation impacts will be low. In headwater streams, planting in low lying valley bottom areas is more likely to be at risk of a high impact of beaver induced flooding. We could use beaver model outputs (Beaver Dam Capacity model) to understand where they are likely to build dams and therefore the potential scale of impact from beaver induced flooding.
  2. Proximity to beaver territory. I can envisage that in some catchments in Devon with low densities of beavers, woodland planting near watercourses will not be impacted by beavers for many years. Here impacts would be none - low. Whereas if planted near an active territory the impacts of foraging and potential beaver induced flooding would be much higher. Again, this will come down to taking into account local beaver population at the time when planting.
  3. Tree species planted. We know that wetland species like alder and willow cope well within wet environments. They are less likely to completely die due to beaver induced flooding. The impacts on woodland planted with these species are more likely to therefore be linked to feeding than beaver induced flooding. However, if trees are planted adjacent to a watercourse that do not cope well in wet conditions then the impact of beaver induced flooding will be high. We have seen the die back of large oaks for example in Devon due to beaver induced flooding (even where the trees have been protected from foraging). Planting blocks of trees less favoured for beaver foraging like alder could be used to create a wider variety of woodland structure, with areas of closed canopy.
  4. The scale of woodland planting and availability of other resources nearby. Large woodland blocks that extend over undulating topography and differ in proximity to the watercourse and beaver territory are likely to have impacts perceived as 'Low or Medium'. Here I can only see the impacts as being positive in terms of creating a wider diversity of habitats. However, small blocks of woodland immediately next to a watercourse where other suitable broadleaf forage habitat is lacking are more likely to have impacts perceived as 'High' if beavers are active in the area. Again using model outputs from the University of Exeter (Beaver Forage Index model) we can estimate proximity to suitable habitat and that would help guide woodland management plans.
  5. I think the complex nature of the likely impacts really highlights the need for clear guidance on what to consider in different situations. It will be brilliant to be kept up to date with the outcome of this work.
• It's simple: get the area and let mother nature do the job. No jobs for biologists, but lots of nature.
• Rewetting and digging activities enable trees and shrubs to settle in otherwise extremely persistent single-species stands, e.g. from reeds or nettles
• Deer browsing can be right through the whole forest at extremes with beavers limited only to the riparian zone.