Railways are pivotal to Europe's ambition for a sustainable transportation system. Yet, the increasing number of animal-train collisions are an overlooked challenge that undermines the reliability of rail services, contradicts environmental policies, incurs socio-economic costs, and results in significant wildlife casualties. These issues are poised to escalate with the rise in train traffic and speed, more vulnerable train designs, and the recovery of wildlife populations. Traditional mitigation strategies, like fencing and wildlife overpasses, are costly, inflexible, and often inapplicable. More adaptable solutions are needed that can meet changing environmental conditions. Several research teams are already studying innovative methods to temporarily deter wildlife from railway areas, thereby reducing collision risks while permitting animals to traverse between train passages. Acoustic warning systems appear here as a suitable tool, but they are only one of several possible and necessary approaches. To develop a comprehensive mitigation strategy on wildlife-railway conflicts, further insights into technological implementations, legal and administrative limits, and wildlife behavior are necessary. The IENE Working Group "Safer Railways for Wildlife” (SR4W) seeks to support this development. During this interactive session, we will showcase the ongoing efforts of the participating research teams in eight lightning talks. Participants are encouraged to engage in discussions on methods, potential synergies and collaborative opportunities. We then engage invited researchers, train operators and stakeholders in a panel discussion to foster a broader, interdisciplinary dialogue, advancing our collective development and implementation of effective, wildlife-friendly railway systems.

The global dependency on railways as an economical and environmentally-friendly option for transportation is steadily increasing. Despite their numerous benefits , railways and train traffic can have negative impacts on wildlife, e.g., mortality due to to collisions with trains or entrapment in rails. In this study, we describe patterns of rail-mortality of the local fauna in Balule Nature Reserve, South Africa. In addition to exploring which species are most vulnerable to rail-mortality, we explore the role that seasonal variation, the daily activity patterns of the species, and the surrounding habitat type have on the occurrence of wildlife-rail-mortality. From May 2020 – March 2021, we surveyed three 5 km railway segments for carcasses. Each 5 km segment of railway was surveyed during the wet season (November–March) and the dry season (May–September) to account for seasonal variation. We recorded 99 rail-kills, of which 29% were mammals, 26% were birds, 26% were reptiles, and 18% were amphibians. Mammal carcasses were found most frequently in the dry season, while amphibians were only detected in the wet season. Amphibian carcasses were all nocturnal species, while diurnal species dominated the bird carcasses found. Finally, most rail-kill carcasses were found in mixed shrublands, while open woodlands had the lowest frequency of rail-kill. The detrimental impacts of railway-mortality on wildlife may be more pronounced in areas of high conservation value, such as nature reserves, and it is essential to study and mitigate these impacts, in order to foster successful co-existence of wildlife and humans in the landscape.

SNCF Réseau, the French railway operator, possesses a network spanning 28,000 km of railroad lines, with some dating back to the mid-19th century. Certain sections traverse wetlands, fragmenting these ecosystems and creating insurmountable barriers for certain species, notably amphibians due to their size which makes it challenging or impossible for them to cross the rails.

Traditional remedies like culverts or small under-track structures face limitations in specific configurations, particularly when the track has minimal clearance above ground level. This is the case of the railroads in the Rhone delta bordering the Camargue regional park in South of France, the designated site for our experimentation.

In response to this challenge, railway experts, alongside scientific and industrial partners, collaborated to devise an innovative solution featuring hollow metal railway sleepers. These sleepers, required to match the dimensions of standard framing sleepers, incorporate various enhancements to increase their appeal and effectiveness. These enhancements include a custom-designed grating to prevent individuals from straying on the track, funnels at the entry and exit points, and deflectors to catch specimens following the rails.

On top of the sleeper characteristics, we will present both the advantages and limitations of this solution, as well as the preliminary positive results observed during on-site evaluation. A comprehensive breakdown of results will be proposed in a poster presentation.

The ecological fragmentation effect of linear transport infrastructures has been widely documented over the last two decades. For railways, this effect is due both to the crossing difficulties for animals (fences, obstacles, fear) and to the collisions risk with trains. Among the levers to reduce ecological fragmentation, the use by animals of crossing structures not dedicated to wildlife has been attested locally by various observations. Only few studies apply ecological continuities concepts to railways, and generally fragmentation and collisions are not addressed together. In this context, a PhD study is being pursued, to better understand the joint influence of various factors on the components of the infrastructures barrier effect (crossing difficulties, mortality), in a multi-level approach (local and regional scale). The frequentation by ungulates – red deer (Cervus elaphus), roe deer (Capreolus capreolus) and wild boar (Sus scrofa) – of 40 overpasses and underpasses not dedicated to wildlife is assessed by video recording using camera traps over a one-year period. Video data processing will be based on automated identification algorithms using artificial intelligence, which will be improved to enhance their accuracy and identify complete and partial passages, along with refusals to cross by animals. Graph-based spatial modeling of ecological networks will be mobilized to assess functional connectivity, which will be investigated as an explaining factor of spatial patterns of collisions and use of crossing structures along a particularly accident-prone railway section in north-western France. The following lighting-talk will address the methodological challenges and ongoing progress of the research.

Enhancing connectivity and minimizing wildlife accidents along railways is a significant challenge for infrastructure authorities. Fences and fauna bridges typically yield positive results, but they are expensive measures, justified mainly on rail lines with elevated speeds and substantial train traffic. Therefore, supplementary measures that can offer wildlife protection along the majority of railway networks are needed.

The Transport Administration applied existing technical equipment for human safety at railroad stations and modified it to a new measure to detect and avert wildlife with sounds prior train arrival at a level fauna passage. Our previous research suggests that the human voice is highly effective in triggering a flight response in ungulates and that ungulates may habituate over time. Hence, we employed an informative voice message solely as a warning signal when ungulates were detected, at the same time as a train approached the site.

The site has been visited at 206 occasions by ungulates during the first year since construction. At 38 of those occasions the warning system was activated. All warning activations led to a behavioral change among ungulates where an increased vigilance (44 %) and delayed flight (55 %) was the dominant response. The absolute majority (92 %) of ungulates stopped moving towards the fauna passage or moved away from the site as the human voice message was activated.

Our studies show that it is likely that a wildlife warning system based on acoustic signals can increase safety for wildlife and train operators at level crossings.

Wildlife-train collisions are a widespread issue, leading to significant costs to society, reduced animal welfare and potential impacts on species' population sizes. Building upon the work of Bhardwaj et al. (2022), this study investigates how various factors influence flight behaviour in fallow deer, roe deer and moose in relation to oncoming trains, as well as wildlife detection from the train driver's perspective. The study uses footage from dashboard cameras mounted on front windshield of trains. Results show that individuals were less likely to flee and increased their FID (Flight Initiation Distance) with increasing distance from the railway tracks. Generally, there was a higher likelihood of flight when the typhoon was used to warn the animals and during dusk/dawn. Flight behaviour of fallow deer was significantly influenced by herd size, with a greater likelihood of flight in smaller groups. In contrast to roe deer and fallow deer, moose were significantly affected by train speed, with flight likelihood and FID decreasing with increasing speed. All species showed a higher probability of flight across track with closer distance to the track and when occurring in smaller groups. Lastly, detection distance was significantly obstructed by vegetation, especially trees along the railroad embankment. Further studies are suggested on a warning system that could be activated as soon as animals are detected by the train driver to increase animals’ FID.  

In Norway, we have recently started a study exploring how animals react to sounds omitted from train-mounted speakers. The objective is to prevent animal-train collisions by using acoustic signals that either warn or temporarily scare animals off the tracks. In our study, we initially rely on the train driver to detect animals on the tracks and to activate a customized camera and speaker system. A specific sound is then emitted to deter the animals away from the track, while the behavioral response is recorded on video. Currently, the omitted sound is randomly chosen from a selection of sounds that have been proven to be effectful in our previous controlled field experiments. We will present the technical approach, the process in the collaboration with the train operators, the lessons learned from installing the speakers, and the problems and possibilities exposed during the development process of this prototype. Multiple factors have been considered, such as the design of the train and the roof, space availability, cabling requirements, access to power sources, stakeholders’ engagement in implementation and setup, and privacy concerns. Our initial experiments are done on freight trains such as Cargo Net's CD321 train, that operates at speeds ranging from 80 to 100 kilometers per hour. Whether the speaker systems are suitable for passenger trains at higher speeds remains yet to be seen.

In the area of rail transport, protection against wildlife accidents is becoming increasingly important.
Therefore, based on the results of the Austrian WiConNET project in cooperation with the Austrian rail infrastructure provider OBB-INFRA, a new electronic system was developed and validated to protect Railways from Wildlife-Collisions. The new system is based on iPTEs “Virtual Fence”, but improved to meet the special requirements of the rail infrastructure:
o Modern trains travel at ever higher speeds. This requires that all units must be activated before a train arrives (electronic fence). Therefore, a wireless connectivity is required.
o Rail tracks are frequently crossing through rural and remote areas, so maintaining the service of the WVC-A system, it needs to be monitored by a remote service centre. This is established via internet connection from the sensors connected via cellular gateway.
Highlights:
The wirelessly connected DD460 Virtual Fence demonstrator system has met the special requirements to reduce wildlife fatalities at rail infrastructure more than 75% within the WiConNET testsites. All elements (sensor-actuator devices, remote triggers and internet gateways) are 100% solar powered and do therefore not need to be connected to a powerline. This enables an easy and cost-efficient rollout of the overall system. Overall costs shall be in the range of 15- 20 k€ per km.
From the WiConNET experience, an improved production-ready version (DD461) will be released in 2024 and deployed together with several rail-infrastructure operators for final validation.
Details of the system will be presented