Wildlife vehicle collisions (WVC) represent a serious threat to road safety. To reduce the magnitude of the problem, various mitigation measures have been proposed. One of the most effective is the use of perimeter fencing, which prevents large species such as ungulates, which cause significant socioeconomic consequences, from accessing roads. However, it is not well understood whether this effectiveness is consistent throughout the year or shows seasonal variations linked to the biological cycles of the species. This study analyzes the temporal patterns of the effectiveness of perimeter fencing on high-capacity roads in Spain, particularly in the case of the roe deer (Capreolus capreolus). For this, a database of approximately 130,000 WVCs that occurred in Spain between 2011 and 2020 was used. The results show that the effectiveness of the perimeter fencing throughout the year for the roe deer contrasts with other ungulate species. This seasonal pattern is explained by the biological cycle of the species, with WVCs on fenced roads concentrating during the pre-reproductive period when young individuals are expelled from their natal group. Spatially, accidents during this period tend to concentrate in areas with secondary accesses and croplands, more open and flat areas, compared to areas with a high proportion of forest and slopes, where accidents concentrate the rest of the year.

Currently, three basic types of wildlife fences are used in Germany: type A, a conventional knot braid deer fence; type B, a chain wire fence; type C, a rigid welded wire mesh fence panels. Since fencing needs to restrict access of many different species with different behaviours, fencing needs to be multifaceted to be effectively. Furthermore, the occurrence of wildlife species with habits such as jumping, climbing or digging determines the effective fence design.
We surveyed road managers in combination with a personal assessment of road sections in Germany and derived recommendations for optimal fencing to reduce WVC, e.g. preventing digging and climbing, application of type B and type C, regular maintenance and construction near the roadside.
In conclusion, wildlife fencing could be a very effective mitigation measure to prevent WVC with common as well as protected species, when proper designed and recommendations of the survey will be consequently considered. Currently many wildlife fences lack functionality due to an inappropriate design and maintenance issues in Germany.

Wildlife-vehicle collisions, involving large ungulates, pose a significant challenge for transport companies around the world, leading to damages, financial, and ecological costs. While various existing solutions, such as fences and visual or olfactory deterrents have limited efficiency, acoustic devices are often presented as promising alternatives. However, regardless of the nature of the deterrent signal (i.e., ultrasonic frequencies, impulsive sounds, or natural bioacoustics recordings), animals quickly habituate and tend to stop responding after a few weeks.
In terrestrial mammals, calls share similarities in their form and communicative functions. For instance, distress and/or alarm calls emitted by highly aroused individuals in threatening contexts are characterized by the presence of harsh sounding component know as nonlinear phenomena. Previous studies have proposed hypotheses suggesting that nonlinear phenomena, such as deterministic chaos, subharmonics, frequency jumps, and sidebands, increase responsiveness and make calls difficult to ignore or to habituate to. However, this hypothesis remains poorly investigated as only one study has shown that meerkats take longer to habituate to calls with subharmonics than tonal calls.
Here, using an innovative method of parametric sound synthesis, we investigate whether the presence of one type of nonlinear phenomena, deterministic chaos, affects habitation in receivers. To do this we experimentally added deterministic chaos to red deer alarm calls and domestic dog puppy whines and showed that it increases conspecific responsiveness and reduces habituation. We conclude that adding nonlinear phenomena to acoustic deterrent signals could increase their effectiveness and durability, helping transport companies to reduce wildlife collisions.

Underpasses have a fundamental potential in mitigating infrastructure barriers for wildlife, considering they are a relatively common feature across traffic networks. Increasing our understanding of the functionality of different design and locality of underpasses is therefore crucial within infrastructure planning and management. We analyzed crossing probability in moose (Alces alces) and roe deer (Capreolus capreolus), common wild ungulates in Scandinavia, in relation to underpasses with different design and locality. We analyzed the design through passage width, height, and length, including different combination of these size dimensions; width*height, width/length, and width*height/length. We also categorized underpasses into three size-levels considering all dimensions. Nearest distance to alternative crossing site was included while controlling for season, group size and whether wildlife visits occurred during daytime or nighttime. We compared 797 visits by moose across 9 underpasses and 1 433 visits by roedeer across 13 underpasses. Increasing tunnel length lowered the crossing probability in both moose and roedeer. Tunnel length not exceeding 13.5 m had a crossing probability above 50% in moose, whereas tunnel length not exceeding 10.0 m had a crossing probability above 50% in roe deer. However, we found no correlation between crossing probability and passage width and height, or other size factor combinations in either moose or roe deer. In roe deer, crossing probability increased when alternative crossing sites was situated further away. The impact on crossing probability from availability of other crossing sites, stress the importance of securing crossing opportunities along infrastructure barriers within a home range scale of focal species.

Cerema was assigned by the Road Department of the French Ministry for Ecological Transition and Territorial Cohesion to take over the data analysis from the collision protocol used by Interdepartemental Roads Directorates (managers of the French national road network, approximately 10,000km, either 1% of French roads but more than one 20% of the national traffic). The term "protocol" should be taken to mean the overall approach to collisions, including field surveys, data analysis and the recommending actions to restore the ecological continuity. The collision protocol has been applied throughout the national road network for almost 10 years in certain/particular sectors/areas. Collisions are recorded daily by patrol officers as part of their road maintenance work. Cerema structured and analysed the centralised data (N=65,500) to identify wildlife-vehicle collision black spots. Within this framework, it has developed a method that takes into account species with similar life traits, the recurrence of collision black spots year after year and the species at stake. Collision blackspots are then prioritized to localize the measures to be implemented to restore ecological continuity, for network managers. At the same time, training courses on the subjects of road ecology and management of green verges are also given to road agents and an animation was also deployed to support agents in the fields survey.

Wildlife roadkill studies need to cope with a mismatch among recorded carcasses and actual road mortality, because of the existence of three biases: crippling, carcass-persistence, and observer bias. Here, we focused on the often overlooked crippling bias, suggesting that it should be called carcass-location bias and disentangling the related three possible outcomes for affected wildlife: injured animal escaping and dying away from road, animal rebounding off the road after vehicle collision, and animal retained by vehicle. Such outcomes can probably be affected by different species traits, and, in order to make a first evaluation of this hypothesis, we opportunistically collected 150 direct observations on the ultimate fate of roadkilled vertebrates. Approximately one third of them were affected by carcass-location bias, so extremely difficult to be recorded through typical roadkill surveys, entailing a considerable and overlooked source of error for roadkill studies and mitigation actions based on them.

The number of wildlife-vehicle collisions (WVCs) has been increasing significantly in recent years in many countries around the world. This negative phenomenon causes significant economic damage and poses a threat to traffic safety, but above all a threat to human and animal life. In Europe, odour repellents (OR) are often used to change the behaviour of wildlife near roads and thus reduce the incidence of WVCs. For two years (in the spring and summer seasons) we monitored roe deer (Capreolus capreolus) using photo traps near roads in the Czech Republic. The behaviour of roe deer (the most common victim of WVCs) was compared between the period without OR and with the OR. Roe deer behaviour was divided into three different phases (arrival, reaction, leaving) and evaluated over time. There was no significant prolongation of the reaction phase after OR application. Individuals refused to cross the road with similar probability with and without the OR.
A negative correlation was found between the time gap of passing vehicles and the reaction time of deer in cases without and with the use of OR. Our results do not indicate a difference in roe deer behaviour without and with OR.
Our findings suggest that OR do not show the expected effect.

Roadkills are both product and symptom of an altered landscape and the spatiotemporal patterns of roadkills are influenced by the existing landscape characteristics, as these determine habitat and resource availability to meet animals requirements. To understand the relationship between roadkill incidences and their habitat corelates, a land use land cover mapping for 2017 was done for the entire highway stretch of 64km and the habitat classes were studied in terms of their frequency of roadkill occurrences. Multi-spectral satellite data was acquired from Copernicus for December 2017. Secondary data included actual ground data, in the form of geolocations collected from January 2017 to December 2017 for image analysis, classification and overall processing of classification results. The geo-coordinates of the carcasses were collected to assess the spatial roadkill pattern and were plotted in the ArcGIS 10.4.1 software. Maximum Likelihood classifier was used for the supervised classification and seven LULC classes were classified. A 500 m buffer on each side of the road was created to assess the surrounding landscape by recording the area of each land cover class. Then the individual roadkill locations for each class of vertebrates, viz. amphibia, reptiles, birds and mammals were incorporated into the generated LULC map. We found that most of the roadkills were found on road sections adjacent to plantation area (40.35%), followed by waterbodies (36.76%) and forested area (7.25%), with least roadkills in sections along grassland (1.61%) and openland (3.02%). Thereby, this study concluded that habitat characteristics made significant contributions toward explaining roadkill.