Wildlife & Traffic
A European Handbook for Identifying Conflicts and Designing Solutions
7 Solutions to reduce transport infrastructure impacts on wildlife
Updated version (2022). Produced in cooperation with the project Horizon 2020 BISON. ‘Biodiversity and infrastructure synergies and opportunities for European transport networks’.
7.2 Reducing wildlife mortality
Transport systems unintentionally kill millions of animals every year. It has been argued that during the past decades, traffic on roads and railways has become one of the major human induced lethal impacts on wildlife on Earth.
At the same time, collisions between large mammals and vehicles or trains cause expensive damage, create a significant additional risk to infrastructure users and are normally fatal for the animals involved.
Small animal mortality may be less an economic and safety problem than an ethical and conservation issue, especially if they are already threatened or endangered. Small animals can be killed by traffic but also get trapped in drains, steep banks or smooth water treatment basins; flying insects can also be attracted to lights where they die or are easily captured by predators. Moreover, these small dead animals on the infrastructure attract scavengers which in turn become vulnerable to traffic.
This chapter presents a variety of measures designed to reduce traffic accidents involving wildlife, and reduce the number of animals killed on or around transport infrastructure, not only by collisions with traffic but also injured or trapped by elements of the infrastructure.
Main types of solutions described in this section are Fencing (7.2.1), Driver warnings (7.2.2) and Wildlife deterrents (7.2.3), that are complemented by Other measures to reduce mortality and disturbance (7.3).
7.2.1 Fencing
General description
Fences, if appropriately designed, installed and maintained, are one of the most effective measures to prevent animals from accessing roads or railways, especially if combined with wildlife passages.
Fences can significantly reduce the risk for vehicle collisions with large animals and, if appropriately designed, could reduce traffic mortality in small animal as well. However, as fencing implies a physical barrier that severely limits the movement of fauna across infrastructure, and is expensive to install and maintain, fencing should only be installed on selected parts of infrastructure where collisions are particularly common.
Fencing, where required, must always be combined with appropriate wildlife passages reducing the barrier effect and providing a safe crossing point for target species (see Section 7.4 – Reducing the barrier effect: wildlife passages). Thus, fences serve two purposes: preventing animals from accessing infrastructure and guiding them to safe crossing points (Figure 7.2.1).
A fencing system must be planned comprehensively and with foresight. It must be based on knowledge about the ecology and behaviour of target species and include:
- Clear identification of target species and specification of mitigation objectives.
- Analysis of landscape features and habitats that need to be reconnected to provide suitable locations for wildlife passages.
- Localisation and design of gates, cattle grids and escape facilities, if required.
- Identification of the necessary length of the fencing system and design of fence endings.
- Decisions about the type of fence mesh, pole type and reinforcements which match the target species’ requirements.
- A maintenance plan.
- Any species that is living in or may move through the area during dispersion or long-distance migrations should be included within the ‘target species’, to inform the location and type of crossing structures and how fences can guide animals to them.

Location and length
- In general, wildlife fences should be installed only where there is a high risk of accidents involving wildlife or hotspots of wildlife mortality. In many countries, perimeter fences are an obligatory safety measure along high-speed railways, motorways and other busy roads.
- On roads with relatively low-traffic intensity fences should be avoided or installed only in sections of particular traffic safety concern or high wildlife mortality. Extensive and unnecessary fencing imposes a barrier that, over time, could be a bigger threat to wildlife populations than traffic mortality. Other solutions to reduce wildlife-vehicle collisions such as driver warning devices and wildlife deterrents may be more appropriate in low traffic roads or railways (see Section 7.2.2 – Driver warnings and Section 7.2.3 – Wildlife deterrents).
- A threshold traffic intensity below which fencing is not recommended must be established according to local conditions and studies. In some Nordic countries fencing is only considered if over 4,000 vehicles per day while in many central and southern European countries the recommended threshold is 10,000 vehicles per day.
- Fences should always be installed on both sides of roads or rail tracks in order to prevent wildlife accessing the infrastructure and being trapped there.
- Fences should always be well attached to the entrances of any transversal structure (at the walls on underpasses or lateral barriers or screens on overpasses, Figure 7.2.2) to prevent animals that approach from accessing the road or rail track. Instead, fences should funnel animals towards the opening of suitable wildlife crossing structures.
- The length of fences should be as short as possible but as long as needed to terminate fences securely, preferably at crossing structures. Fence ends must be given special attention as they are potential entry points for wildlife to access the causeway.
- Fence openings, for connecting local access roads or trails, can provide unwanted and dangerous entry points for wildlife. They must be secured to prevent animals from entering the fenced causeway. This can be achieved by gates, cattle grids or extending the fence along the minor road.
- Fences will never guarantee complete traffic safety. Occasionally, animals will succeed in entering the causeway and be trapped in the fenced corridor. These animals are a severe safety hazard as they may behave erratically and unpredictably and because drivers do not expect animals within the secured causeway. So, the installation of escape facilities (jump out, one-way gates or other) could be necessary near openings or fence ends (see Section 7.2.1 – Escape devices)

- To avoid;
- Less favourable;
- Optimal.
- At interchanges, specific solutions must be defined according to site features. Fencing may continue along access and exit routes to reduce the possibility of wildlife entering infrastructure (see an example at Figure 7.2.3A). Alternatively, fencing could be limited to the main infrastructure, with openings only for access and exit routes while local roads with low traffic may be adapted, combining road and wildlife crossings at both sides. Installing cattle grids at access and exit routes could also help to prevent wildlife accessing the fenced road (Figure 7.2.3B).

- In order to ensure the maximum available habitats and lateral corridors for wildlife adjacent to infrastructure, fencing should be placed as close to the infrastructure as possible, considering traffic safety and maintenance issues (Figure 7.2.4). This proximity provides more favourable habitats on the exterior side and can discourage wildlife from trying to reach habitat areas inside the fences. Appropriately maintaining the vegetation on road verges inside the fenced area (see Chapter 10 – Maintenance of fencing: meshes and poles and Chapter 10 – Management of vegetation) could reduce the attraction of animals to the carriageway, diminishing the mortality risk. The positioning of the fence as close as possible to the infrastructure may lead to difficulties in the management of areas outside the fence for operators. Gates in the fencing or exterior access points could remedy this problem. It is also possible to transfer management to external operators (for example by agreements with farmers or other local stakeholders). Stone blocks or similar may be used to mark public property boundaries.

- When fences are installed in combination with underpasses, the optimal position is at the top of the embankment with road verges to the exterior (Figure 7.2.5A) and providing safe guidance towards the underpass entrance. When a fence can only be placed at the base of the embankment, it should funnel wildlife towards the entrance (Figure 7.2.5B). Fence ends should be securely attached to the crossing structure entrance. Fences must not be positioned perpendicular to the causeway because animals may struggle to find the entrance to the crossing structure (see Figure 7.2.5C) and this design may create a tunnel-effect that deters some animals from accessing it.

- If fencing is not continuous, but installed only in conflictive sections, each fence end must be connected to the walls of any potential crossing structure (culvert, overpass, underpass, viaduct, etc.) that could be adapted to be a safe multiuse crossing for the target species. Where fencing cannot be connected to any crossing structures, the point at which the fence finishes could result in a hot spot of animal-vehicle collisions because some species, such as deer or wild boar, will follow the fence even for long distances in search of a crossing point. A GIS analysis which evaluates how much adjacent habitat is suitable for target species, the location of road mortality hotspots and other features of the ecological corridor and networks is suggested when assessing the position and length of a fence. A standard length of 500m both sides of crossing structures has been proposed but much longer distances have been suggested as necessary for large mammals (several kilometres). See Rationale box. Fencing design
- Where non-continuous fencing must be installed and infrastructure is at level (under or overpasses are not available) it may be necessary to consider the installation of Animal Detection Systems which activate warning sign to drivers when large animals are present or even deter animals from crossing when a vehicle is approaching (see Section 7.2.2 – Warning signs activated by animal detection systems (ADS) and Section 7.4.4 – At grade fauna passages)
- Where concrete safety barriers are located in the medians or at the borders of an infrastructure a perimeter fence is required on both sides of the road to avoid animals accessing and becoming trapped on the causeway or taking longer to cross the road. The recommended location for fencing and safety barriers is shown in Figure 7.2.6.

Design and dimensions
RATIONALE. Fencing design
There is a lack of published information regarding the effectiveness of the different types of fencing, mesh types, height and length (Rytwinski et al., 2016). The technical prescriptions regarding design and dimensions included in this present chapter are mainly based on knowledge and experience from infrastructure operators and experts from different countries (AT, Carpathian region, ES, FR, NL and SE) and guidelines and handbooks providing best practice (van der Ree et al., 2015; Animex, 2021).
When fencing is installed only on particular sections of the infrastructure, to define the appropriate fence length to prevent roadkill is a particularly difficult decision. Any fence should be long enough to avoid or reduce the ‘fence-end effect’ (Clevenger et al., 2001; Spanowicz et al., 2020), which occurs when animals such as ungulates follow fencing and cross the infrastructure at the point it ends. To avoid this effect, fence ends must always be connected to crossing structures. When this is not feasible, fence ends must be placed in areas where animals do not tend to cross (Clevenger and Huijser, 2011) or in areas with high visibility to drivers (Barthelmess et al., 2014).
Jaeger & Re (2021) developed models to assess appropriate length according to target species’ home range and location regarding fencing where a percentage of roadkill reduction (effectiveness) is calculated as %E = 1 – 0.226 * (R/L), where R is the radius of the target species home range and L is the length of the fence. A similar approach was developed by Seiler et al. (2015) to determine a standard for the maximum length of a fence to avoid barrier effect for ungulates in Sweden, determining that in the study area fences longer than 6 km require combination with crossing structures.
- Conventional wildlife fences consist of wire mesh fixed to regularly-spaced poles that should be buried into the ground or appropriately anchored to avoid wildlife passing under the mesh (Figure 7.2.7). The basic model could be complemented with reinforcements to avoid access by specific target species, such as amphibians, large mammals, climbers or burrowing species (see Section 7.2.1 – Fencing and reinforcements for small vertebrates and Section 7.2.1 – Adaptation of existing fences and other systems to deter burrowing animals)

- The characteristics of the fence (i.e., height, mesh type) must be defined according to target species and following these criteria:
- Fence is of sufficient height to discourage large mammals from attempting to jump over it (Figure 7.2.8).
- Mesh (or any other material to be used) is of the correct type, size and condition to prevent target species passing through or climbing over it. The mesh is durable and reduces the risk of entangling and injuring animals (Figure 7.2.9; Figure 7.2.10).
- Mesh is appropriately buried in or anchored to the ground to ensure animals cannot access the infrastructure by going underneath it.
- Fence fixings and base anchoring are properly installed or the base of the fence buried to prevent animals passing under the fence (see Section 7.2.1 – Poles and Section 7.2.1 – Adaptation of existing fences and other systems to deter burrowing animals)
- Wildlife often try to cross infrastructure looking for more suitable habitats. Ecological restoration of target species’ habitats to attract them to particular sites related to the infrastructure could be part of the fencing strategy. These quality habitats could play a role in reducing the need for fencing or helping to reduce the cost of bigger and more robust fences because the animals don’t attempt to cross as frequently.
- Recommended fencing, including mesh type and other materials to be used and dimensions according to target species, is shown in Table 7.2.1.
Table 7.2.1 – Dimensions and type of fences suitable for different species or groups of species according to standards in different European countries (see Rationale box. Fencing design. Fencing design and dimensions should be adapted to target species and site conditions (Figure 7.2.11).
- The height of the fence should be determined by the largest mammals likely to attempt to jump over it (Table 7.2.1) and by the position of the fence in relation to the surrounding terrain. This height must be constant and measured on the approach side for the animals – the side away from the infrastructure – so the adjacent terrain must be considered (Figure 7.2.8).
- The height of snow in winter, ‘health and safety’ regulations for maintenance operators and other factors must also be taken into account when defining the height of the fence.

Fence material and mesh types
- The type of fence to be installed should take account of main target species to be excluded from the infrastructure and must be designed accordingly.
- The mesh, or any other fence material, should be fixed on the outside of the poles (i.e., away from the road) to prevent a large animal being able to bring down the mesh by pushing its attachments to breaking point.
- Wire mesh is the standard material applied in wildlife fencing where large mammals are the target. Three types of wire mesh are often used (Figure 7.2.9): knotted mesh (regular or progressive density), chain-link mesh and welded mesh. Several dimensions of mesh and wires are provided by manufacturers. Knotted mesh is the standard for large mammals but it could pose a risk for felines with reports of specific cases of wildcats being trapped in this type of wire mesh. Chain-link is not suitable where wild boar is a concern because this species can deform it and open holes. Welded mesh is safe for all types of fauna, and is particularly strong and suitable in areas requiring maximum resistance, but it is more costly. Balancing efficiency and cost is crucial in choosing the most suitable mesh type.

- Other materials such as steel or concrete are also recommended to deter small fauna -particularly climbing species- and to avoid harming them (Figure 7.2.10). Materials that do not have long durability or produce high amounts of plastic waste should be avoided. Nevertheless, polymers of recycled plastic which have been guaranteed to be durable may be suitable for small fencing reinforcements (see Section 7.2.1 – Fencing and reinforcements for small vertebrates and Section 7.2.1 – Adaptation of existing fences and other systems to deter burrowing animals)

- Fences must be strong and durable enough to resist attempts by wildlife to break through. A rectangular knotted mesh fence with a wire made of stainless materials (galvanized) and with a diameter of at least 2.5 mm is the standard wildlife fence recommended for large mammals (ungulates and large carnivore). In areas with heavy snowfall, the top netting wire must be reinforced with a cable capable of bearing the weight of snow settling on it.
- The dimension of the mesh (Table 7.2.1) is determined by the size of the animals it should stop and their ability to squeeze through it, not forgetting juvenile animals which may end up getting through. Standard wire mesh density is generally 15×15 cm, because ungulates are the largest target mammals in many areas.
- Knotted wire mesh of progressive density (wider at the top) is recommended for large mammals, particularly ungulates, and can also prevent medium-sized mammals like foxes or badgers from crossing (Figure 7.2.7).
- The lower part of the mesh should be buried in the ground to combat digging animals (such as wild boar, badger or fox). Mesh can also be extended horizontally to prevent brown bear getting through. Alternatively, ground pegs have been recommended to fix the mesh to the ground, even if the effectiveness of this measure depends on soil hardness and other factors. Welded mesh reinforcements could be installed when burrowing animals overcome existing fencing (see Section 7.2.1 – Fencing and reinforcements for small vertebrates and Section 7.2.1 – Adaptation of existing fences and other systems to deter burrowing animals)
- Different types of fences could be combined to adapt fencing to target species requirements. The combinations most frequently applied for large mammals in Europe are shown in Figure 7.2.11 and examples in Figure 7.2.12. Note that these are not suitable for amphibians and reptiles.


- Modular fencing is being applied in some countries, consisting of different panels that could be assembled according to target species in the area (Figure 7.2.13). They could integrate 1) outrigger panels preventing access by climbing species (mesh, metal sheets, or even floppy overhangs in specific cases); 2) chain-link, knotted or standard mesh panels, and 3) opaque buried panels to prevent access by burrowing species as well as amphibians and reptiles.

Fencing and reinforcements for small vertebrates
- To avoid small animals (hedgehogs, mustelids, amphibians, reptiles etc.) accessing the infrastructure at sections where it crosses particularly vulnerable habitats, additional reinforcements must be installed at the lower exterior part of a standard fence for large mammals. These fences can also be installed alone (Figure 7.2.14).
- The entire fenced length of the infrastructure does not necessarily have to be impenetrable for small fauna, because in some sections the verges can be used by these animals as habitats for living or moving.
- Fence sections with specific additional precautions to prevent small wildlife access are generally to be reserved for:
– Intersection of ecological corridors that wildlife use to move across the landscape and where safe crossing structures must be provided to prevent mortality.
– Habitats of high risk which are ‘ecological traps’ attracting wildlife to areas where they will experience high mortality and which could even jeopardise the long-term viability of a population.
– Road sections where median safety concrete barriers are installed (Figure 7.2.6) which are potential traps for small and medium-sized animals.
- To provide fencing which is effective and safe for amphibians, reptiles and other small fauna, smooth opaque and durable materials should be used, such as metal, concrete or recyclable resistant polymers (Figure 7.2.10 and Figure 7.2.14). These fences could be added as a reinforcement to standard fencing or be used alone, fixed on small poles. Some manufacturers provide small fauna fencing which is smooth at the exterior part to avoid climbing but rough on the interior, allowing the small fauna trapped in fenced sections to escape to adjacent habitats. Combinations of amphibian fencing which guides the animals to suitable crossing points are described in Section 7.4.3 – Amphibian passages

- A high-density welded mesh anchored to the lower part of a fence for large mammals is a frequent practice to prevent access by hedgehogs, badgers or other small and medium size animals. Reinforcements must be installed at the exterior side of the standard fence. Height (40 cm to 1m) and mesh density size (3×3 or 4×4) must be defined according to the species to be stopped (Table 7.2.1). Mesh fences are not suitable for amphibians and reptiles because they can easily climb over mesh fences (Figure 7.2.15). Amphibians may also be harmed if they get trapped in the mesh.

- To avoid;
- Less favourable;
- Optimal.
Adaptation of existing fences and other systems to deter burrowing animals
- Where animals, particularly brown bear, wild boar, badger, foxes or rabbits, are accessing infrastructure under existing fences, reinforcements can be installed to make these more robust (Figure 7.2.16). Reinforcements could also be included in the design of new fencing in sections where conflicts with burrowing species are expected.
- Several materials can be used to deter burrowers. Panels of metal or recyclable durable polymers could be applied such as in the case of small fauna (Figure 7.2.13). One of the most widespread practices in Europe is the use of welded wire mesh panels that are attached to the exterior part of a standard fence for large mammals and buried into the ground.
- Dimensions and design of the welded reinforcement mesh are according to target species (Table 7.2.1). This will most often require:
– Burying the reinforcement fence 20 to 50 cm into the ground.
– In the case of any persistent burrowing by smaller species such as fox or wild boar burying the reinforcement fence in the form of an L towards the outside, to a depth of 20 cm, with a 20 cm horizontal fold outwards has been proved to be effective (Figure 7.2.17).
– Where wild boar is the target species, an effective solution is to install additional welded mesh 30×5 cm buried 20 to 50 cm at the bottom of the fence. Standard panels that can be nailed to the soil are available. Long pins driven into the ground to hold the bottom of the fence to the ground could be an alternative method to fix the mesh. Despite not being particularly effective, this could provide a solution where the ground is hard.
– Where brown bear is the target species, an additional wire mesh can be installed at the bottom and buried in a L-shape design: 20 cm vertical and 120 cm horizontal and covered by soil (Figure 7.2.18).


- The standard large mammal fence could be complemented with modular panels to prevent climbing and digging, or with opaque guiding fences for amphibians and other small fauna (Figure 7.2.13).
- Electrified wires and reinforced wire mesh could also be used in particularly conflictive fenced areas. Technical prescriptions vary according to target species and national regulations must be considered before applying such solutions.


Poles
- Fences must be fixed onto sufficiently solid poles, which can be made of galvanized steel (an alloy of 95% Zinc and 5% Aluminum is suitable) rot-proof wood, or even concrete. Steel poles are generally T-shaped or hollow with a square, rectangular or tubular section. The specific characteristics of the poles must take into account both the solidity of the structure and, in certain cases, safety aspects to prevent the fence from adding to the severity of a traffic accident. The poles can be anchored in two ways (Figure 7.2.20):
– Grounded in a concrete block (depth of 40 to 70 cm depending on the nature of the ground)
– Driven directly into the ground by hammering (approximately 70 cm or more depending on the soil). Metal anchor pegs can also ensure the stability of the fencing while limiting the need for concrete blocks.
- The distance between posts must be 4 to 6 m for large fauna except brown bear and wild boar, which require a lesser distance of 2 to 4 m. Terrain features may also dictate a shorter distance between posts.
- Suitable anchorages for tension posts (maximum distance of 5 m between two tension posts) are required.

Escape devices
- Even with the best fences, infrastructure is never completely sealed off to wildlife. Occasionally, animals may access carriageways at road interchanges, gates, sections damaged due to e.g. landslides, fallen trees, maintenance machinery or even vandalism. It is thus advisable to install escape facilities where required to permit the animals exit at sections where access is more likely.
- Escape facilities are one-way devices that allow wildlife to exit the infrastructure while preventing their re-entry. Their frequency and positioning must be adapted to the features of the area and the configuration of the fence. Two important considerations are:
– Adequate maintenance must be guaranteed because these devices may fail and become entrances for animals.
– Even if installation of these devices is a usual practice in some regions or countries, there is a lack of evidence about their effectiveness and more research is required. Therefore, escape devices should only be used in places where they are thought to be absolutely necessary and a monitoring programme must be planned.
– Different types of escape systems may be used according to particular species. Main types are escape ramps, jump-outs and gates.
- Escape ramps on the infrastructure side of the fence which form a gently sloping mound up to almost the top of the fence (Figure 7.2.21). Animals trapped in the infrastructure tend to follow fences and can climb these mounds and jump down to the other side, from where they cannot get back in. The design and positioning of the device must not pose any risk in the event of a traffic accident. This type of escape device is particularly intended for large ungulates, but the same principle is used for carnivores and even small vertebrates (Figure 7.2.23; Figure 7.2.24). There is a lack of knowledge about the height that different European target species can jump with no risk of injuries. Fences must be well connected to the sides of the jumping ramp or be continuous (Figure 7.2.21; Figure 7.2.23).
- Jump-outs are another type of escape device which is particularly recommended for deer (Figure 7.2.22). The animals walk along the fence until they find these devices that allow them to jump out of the fenced area. Usually they are placed at the base of embankments. The same concerns mentioned for escape ramps apply here.

- To avoid;
- Less favourable;
- Optimal.
- To avoid;
- Less favourable;
- Optimal.


- One-way gates open only in one direction to the outside of the infrastructure and close automatically after the animal passes through, either by weight or a tensioned closing mechanism (Figure 7.2.25). The size depends on the target species and a wide variety of designs exist. While their effectiveness is not clearly demonstrated, these systems are often installed for medium-sized animals such as badgers, fox or wild boar, but also for ungulates (Figure 7.2.25B). A particularly careful maintenance schedule must be provided for these devices as they can eventually seize up, get stuck open and are at risk of doing the opposite of what was intended by creating an easy entry point for wildlife to the fenced area. This exit device also requires a hard base to prevent animals digging under it and vegetation growing that blocks its proper operation.

- To avoid;
- Less favourable;
- Optimal.
- The effectiveness of the escape facilities depends on regular maintenance (see Chapter 10 – Maintenance of fencing: escape devices). This includes cleaning and repairing natural damage such as a fallen tree or earth movement, wildlife attempts to create access points, such as digging under the fence and also unintentional damage like a traffic accident or impacts from maintenance equipment. Access to maintain the fence system must be provided. Vegetation maintenance is important in order to prevent branches from weighing down the top of the fence. It may be necessary to maintain a strip free from trees and bushes along the fence with a clearance of about 1 m to 1.50 m. Depending on the situation, this should take place on at least one side of the fence.
Cattle grids and gates
- Gates could be installed on small roads with sporadic use, e.g. service roads, and those which provide access to a fenced road. The possibility of entry by small animals under the base of the gate should be limited using a concrete threshold accompanied by a rubber base skirt (Figure 7.2.26).

- To avoid;
- Less favourable;
- Optimal.
Cattle grids are an alternative to gates (Figure 7.2.27). They are a 2 to 3 m wide device consisting of a pit at least 30 to 50 cm deep covered with a metal structure consisting of a series of steel bars spaced about 5 cm apart. No gaps between fencing and the grid should be left. The pit should allow all small animals that fall into it to escape. This can be achieved by sloped lateral walls (Figure 7.2.28), lateral openings or ramps angled at 30-45º (see Chapter 10 – Maintenance of cattle grids). National traffic regulations should be checked as installation is forbidden in some particular conditions and countries. In areas with longer duration snow cover, cattle grids are not suitable as animals are at risk of harm through slippage when attempting to cross.


Other points of attention
- Fences must be thoroughly checked as part of the ordinary road inspection schedule at least once a year and more frequently during the first year after installation (see Chapter 10 – Maintenance of fencing: meshes and poles). Beyond inspection for regular maintenance, particular attention must be paid to how mesh is fixed to the posts, to connections with the various devices, and how the fence is fixed to the ground (Figure 7.2.29).

- To avoid;
- Less favourable;
- Optimal.
- Planting vegetation beside fencing is a controversial topic which requires monitoring and evaluation to provide any evidence-based recommendation. A dense row of bushes planted along the fence exterior is recommended in some European countries with the aim of discouraging deer from jumping over and to reinforce the fence corridor effect, guiding wildlife to crossing structures. On the contrary, other countries advise against this practice because it has been observed that some species such as roe deer, attempt to jump over vegetation with the risk of striking or breaching the fence (Figure 7.2.30A).
- A corridor free from bushes and tree branches beside fencing is recommended by many transport operators. The main aim is to avoid damage caused by vegetation to the fence. It also could prevent species such as pine or stone marten, from climbing over the fence (Figure 7.2.30B).
- In every situation, the planting of vegetation that provide edible fruit or is attractive to foraging animals must be avoided.

- To avoid;
- Less favourable;
- Optimal.
- Special attention should be paid to places where fencing crosses a ditch. Devices can be installed to allow water to flow while preventing wildlife from entering infrastructure (Figure 7.2.31).

Fencing and screening for flying fauna
- Fencing and screens can be used to avoid bird and bat collisions with vehicles, either by increasing the flight height over the infrastructure or by guiding them to wildlife crossing structures. In both cases, the effectiveness of this mitigation measure will depend on species-specific characteristics
- Fences that raise the flying altitude of bats may be more effective in species that are adapted to open areas, which usually fly at higher altitudes and are less manoeuvrable. These fences also seem to be effective for pipistrelle bats that, despite flying 4-5 m above ground, tend to keep a flight height when crossing roads with screen on both sides. Other bat species adapted to the forest or forest edges usually return to their low flight altitude once over the fence, crossing the infrastructure at a height with more risk of collision with vehicle. For these species, the use of fences and screens to divert them from high risk areas and guide them to the crossing devices over or under the infrastructure are necessary (Figure 7.2.32). To increase the effectiveness of this mitigation measure, several characteristics must be considered:
– The guiding fence or screen needs to be of sufficient height (4-5 m) (Figure 7.2.33). In the case of an underpass, an additional fence or screen can be placed above the crossing structure to help guide animals within it. In these cases, the height of the fence or screen can be lower, ca. 2 m (Figure 7.2.2B).
– The size of mesh in any fence must be small (2×2 to 4×4 cm).
– The guiding structure must extend for at least 50 m on either side of the crossing structure.
– Fencing to guide bird or bat flight must be appropriately anchored to the entrances of the crossing structures to avoid gaps though which animals could access the road or railway (Figure 7.2.32).
– Vegetation can be planted to complement the guiding structure, but is of limited success where there has been an attempt to modify well established commuting routes to suit the structure. Therefore, it is better to place the crossing structure where known commuting routes exist (see Section 7.4.2 – Bat crossings).

- To avoid;
- Less favourable;
- Optimal.

- To avoid bird collision with wire meshes, white marks fixed at the upper part of the mesh may help to make fencing visible for birds (Figure 7.2.34). This is particularly recommended for species with slow and winding flight patterns.

7.2.2 Driver warnings
Warning signs aim to influence driving behaviour, reduce speed and increase attention, thus resulting in a reduction in the risk and severity of animal-vehicle collisions (AVC). Several studies have shown that vehicle speed and drivers’ attention are two important factors in wildlife-vehicle collisions. A reduction in speed provides more reaction time, braking distance and reduces the energy of an impact.
Further studies are required to establish the effectiveness of many devices described in this section.
Standard and temporary wildlife traffic signage
RATIONALE. Wildlife warning signs
Wildlife traffic signs are one of the most used mitigation measures, aiming to alert drivers about the potential presence of wild animals on the roads and thereby reducing the risk for AVC. A literature review of the topic is provided by Huijser et al. (2015). Standard (static) wildlife warning signs are more commonly used, although they are generally less effective as drivers become habituated and tend to ignore them (Found & Boyce, 2011; Bullock et al., 2011; Huijser et al., 2015; van der Grift et al., 2017). Enhanced wildlife warning signs, which are bigger, transmit more specific messages, or include eye-catching details, may initially be more noticeable to drivers, but suffer from the same habituation effect as standard signs (Huijser et al., 2015).
Temporary wildlife warning signs are more specific as to where and when AVC risks are elevated, which makes them more reliable for drivers and hence more effective (Sullivan et al., 2004; Gagnon et al., 2010; Huijser et al., 2015). Active warning signs triggered by Animal Detection Systems (ADS) display even more acute information about the immediate accident risk. Active devices are more effective than temporary and standard warning signs, especially if combined with temporary speed limitations.
Both static, temporary and active warning signs can be applied to roads with low-intermediate traffic volumes and low-intermediate speeds, where drivers are able to reduce speed suddenly or even stop vehicles to avoid animals. On motorways with high speed and busy traffic, wildlife warning signs may be pointless or even hazardous. Here, only the physical separation of wildlife and traffic by fences and wildlife passages is recommended from a traffic safety perspective (Huijser et al., 2015; van der Grift et al., 2017).
- Standard wildlife traffic signage is placed in areas where collisions often occur and consequences for human safety are significant. Studies indicate that there is no corresponding reduction in accidents involving wildlife, so these standard signs do little more than provide a legal protection to road operators (see Rationale box. Wildlife warning signs) Road users quickly get used to such signs, especially if they are present all year round and if the driver uses that stretch of road frequently (Figure 7.2.35A). Drivers ignore permanent signs probably because the sign doesn’t often correspond with real presence of an animal close to the road.
- The combination of a wildlife warning sign with a speed limit is slightly more effective. The effectiveness is further enhanced if signs are marked with flashing lights or a flashing speed limit sign, lit only during periods of high target species activity.
- Temporary warning signs, activated only during critical periods of the year, in locations with high risk of collision, have shown higher effectiveness (in terms of AVC reduction) than permanently installed signs. These temporary signs could be reinforced with flashing lights or reflective panels (Figure 7.2.35B).

Warning signs activated by animal detection systems (ADS)
Warning signs associated with animal detection systems (ADS) are devices which include sensors to detect animals of different size and activate warning signs to drivers. ADS may also trigger deterrents to scare away animals when vehicles approach and are described in Section 7.2.3 – Wildlife deterrents. They can only be used on relatively short road sections such as hotspots of animal-vehicle collisions or fence openings.
- ADS require very particular road verge features and must be frequently checked and maintained, as many factors such as growing vegetation can affect their functionality (see Chapter 10 – Maintenance of signs activated by Animal Detection Systems (ADS)).
- ADS may rely on either active light barriers, passive infrared sensors, thermal cameras or radars to detect wildlife movements close to infrastructure (Figure 7.2.36). The power supply is provided by a battery and a small built-in solar module.
- Where ADS are installed they activate warning signs or panels informing drivers that when a flashing light or message appears it means an immediate collision risk because an animal is close to the road and could cross at any moment. It is recommended to combine the warning with a speed reduction sign.
- Wildlife warning devices with ADS are being improved to reduce false positives (activations not caused by the presence of an animal) with a corresponding increase in reliability.

- Animal detection systems can also be combined with a doppler radar that detects approaching vehicles and their speed to trigger light or noise deterrents that are intended to keep animals off the road. A driver warning may be activated when an animal is detected, and if the speed of the approaching vehicle is higher than a given threshold. Messages such as ‘Animal on the Road’ or ‘Slow down’ could be activated. Optionally they can be fitted with animal deterring sounds or signals (Figure 7.2.37). Similar systems are also being tested to deter ungulates on train lines. Studies have demonstrated an increased effectiveness when sounds are used that animals associate with a real danger, such as human voices or dogs barking (see Section 7.2.3 – Wildlife deterrents)

Anti-collision systems in vehicles
- Some car manufacturers have already installed systems in their vehicles to detect people and animals on the road, which assist in braking and speed control, and prevent accidents. Such intelligent technologies are quickly gaining a foothold as essential components in the development of automated and autonomous vehicles (Figure 7.2.38).
- Anti-collision systems may use radar or passive and active infrared cameras to detect potential upcoming collision risks. Algorithms can be employed for image recognition and differentiation of species as well as appropriate response behaviour. Combined with automated braking systems that outperform the driver’s own reaction, collision risks in traffic are likely to be significantly reduced.
- New technologies will probably enable vehicles to communicate with others and share information about the presence of potentially hazardous wildlife near the roadway. Early warnings by other vehicles will help to prepare Artificial Intelligence systems and drivers to react in the most appropriate way. It is however still unclear how well these vehicle based multi-sensor prevention systems may perform and whether they could replace infrastructure based installations. Most likely in-vehicle systems will complement safety measures on minor roads where physical mitigation such as fencing, crossing structures may not be recommended. Whether such systems also would be useful in preventing collisions with small animals is unknown.

7.2.3 Wildlife deterrents
RATIONALE. Deterrents to wildlife
Wildlife deterrents in general are intended to cause fear or discomfort to animals and thereby induce a flight response or at least increased alertness. Various acoustic, visual and olfactory signals can be deployed to achieve this. Available evidence to date shows greater effectiveness of acoustic deterrents that rely on natural sounds over artificial, visual or olfactory cues. The challenge is to develop signals that are readily perceived and understood by wildlife.
Visual deterrents that either reflect or emit light when car headlights shine on them, have been installed widely despite a lack of evidence about effectiveness. Several reviews revealed that reflectors have little or no effect in reducing wildlife-vehicle collisions (D’Angelo & van der Ree, 2015; Brieger et al., 2016, 2017; Kämmerle et al., 2017; Benten et al., 2018a, b, 2019). These authors also stated that the effectiveness previously reported is mainly due to inappropriate study design, leading to potentially unreliable conclusions. Other reasons for this ineffectiveness are e.g., that the animals are unable to perceive the light being reflected (D’Angelo & van der Ree, 2015; Benten et al., 2019; Coulson & Bender, 2019), that the light is insufficiently intense and masked by the already illuminated road surface or roadside vegetation (Seiler et al., 2017), that animals can run onto the road to escape and also that they quickly become habituated (Ujvári et al., 2004; D’Angelo et al., 2006; Coulson & Bender, 2019). One major concern about visual cues is the lack of animal response to the standard headlights of vehicles that illuminate the road ahead.
Innovative acoustic deterrents are showing more promising results, especially those that use sounds with natural meaning, such as alarm calls from local species or human voices (Biedenweg et al., 2011; Babińska-Werka et al., 2015; Seiler & Olson, 2017; Coulson & Bender, 2019; Berndt, 2021; Lodnert, 2021 and references therein), and those that sound only when an actual threat is present in the form of an approaching vehicle (Babińska-Werka et al., 2015; Seiler & Olson, 2017; Lodnert, 2021). These features can improve the results also in the long-term by reducing or eliminating wildlife habituation, which is otherwise the main drawback of deterrents (Elmeros et al., 2011; Benten et al., 2018b).
Acoustic and visual stimuli may also be combined to produce a ‘virtual fence’ that are intended to deter animals from crossing roads or railways when vehicles approach at night. Evidence for the effectiveness of these combined approaches is still scarce and contradictory (Coulson & Bender, 2019; Englefield et al., 2019; Fox et al., 2019; Stannard et al., 2021) and more research is needed to determine their potential.
In general, studies on the sensory perception and flight behavior of the target species should be conducted prior to extensive deployment of mitigation strategies (D’Angelo et al., 2006). Also, more knowledge is needed about the potential disturbance of wildlife deterrents to adjacent habitats.
Visual, acoustic, or olfactory devices, are deployed to help prevent wildlife, often deer and wild boar, from crossing transport infrastructures or to increase their alertness towards approaching vehicles. Development is very active in this field and many types of devices are available. However, so far, most methods have failed to provide convincing evidence of their effectiveness, while some (e.g., wildlife reflectors or vehicle-mounted whistles) are proven to be ineffective.
Some studies that show a reduction in the number of accidents are short-term and do not necessarily prove the effectiveness of such systems. Other reasons can also lead to a reduction in accidents, such as a decrease in animal population or a change in traffic volume.
- Maintenance requirements are a determining factor as many of these devices require high maintenance and do not work if this is neglected (see Chapter 10 – Maintenance of signs activated by Animal Detection Systems (ADS)
- Wild animals, and particularly ungulates and large carnivore, may get used to these stimuli when they are recurrent and not associated with a perceived risk for the animal. This results in a loss of effectiveness over time. To reduce the risk of habituation, these prevention measures may be best deployed only during critical periods when most accidents occur or, preferably, only when vehicles approach.
- Wildlife deterrents could also cause disturbance that may reduce the quality of adjacent habitats next to infrastructure.
Visual deterrents
- Visual deterrents reflect or emit light, normally red, blue or white (Figure 7.2.39) and they are typically attached to delineators located at road verges.
- Reflectors which bounce the lights of approaching vehicles into the surroundings are intended to alert animals, discouraging them from entering the road or scare them away. However, studies show that such lights are not perceived by target species or are too weak compared to the simultaneously illuminated road surface, side vegetation and delineator pole itself. The effectiveness of such devices must therefore be questioned (see Rationale box. Deterrents to wildlife)
- No long-term studies have so far been able to show a sustainable reduction in wildlife accidents where reflectors are deployed. In some cases, a reduction in animal vehicle accidents has been attributed not to any change in animal behaviour but to drivers cutting speed when they see reflected or emitted light, particularly blue, on the road verge. However, other studies do not show any reduction in speed.

- To avoid;
- Less favourable;
- Optimal.
Acoustic deterrents
- Currently available devices that incorporate acoustic deterrents are often attached to delineator poles and triggered by the light of an approaching vehicle or the presence of large animals detected by thermal cameras, movement sensors, or other systems. The devices emit an artificial whistling sound in a frequency audible to the animals that aims to discourage wildlife, mainly deer, from crossing the road or railway. However, there is yet no convincing evidence for the efficacy of these devices in reducing accident risks.
- Similarly, so-called deer whistles, which are mounted on the vehicle and produce sound via the airflow are intended to scare deer away so that no collision occurs. However, the sound may be ineffective because it is outside the audible range of the animal or is drowned out by traffic noise.
- Combined acoustic and visual deterrents, also known as ‘virtual fences’, are also increasingly installed. Attempts are made to reduce the habituation effect by making it possible to vary the sound frequency or colour of the light (Figure 7.2.40). Studies show contradictory results on effectiveness (see Rationale box. Deterrents to wildlife)
- Bio-acoustic approaches that, instead of a bell, chime or whistling sound, use a natural sound that animals associate with a real threat, such as human voices, dogs barking or a species-specific alarm show more promising results. Several studies show a strong flight response in deer and wild boar when human voices are emitted, but the effect may be context dependent. More studies, especially long-term monitoring, are needed to further develop and refine bio-acoustic systems to prevent animal-vehicle collisions.
- Another method which has been proposed is a grooved or rippled road surface that produces a loud warning sound when vehicle wheels move over it. These sounds must be in the frequency range of target species for this to work. However, more detailed investigations are needed to verify its effectiveness and concerns about impact of such a noise on the adjacent habitats have been raised.

- To avoid;
- Less favourable;
- Optimal.
Olfactory deterrents
- Olfactory deterrents or so-called ‘scented fences’ are intended to keep wildlife away from crops and urban green spaces and they are also proposed as a method to prevent accidents on infrastructure (Figure 7.2.41). Synthetically produced fragrances which imitate human or predator scents are often deployed. These products should be used only during short critical periods because animals become readily habituated (see Rationale box. Deterrents to wildlife)

- To avoid;
- Less favourable;
- Optimal.