Recently, there has been a rapid rise in the frequency of wildlife-vehicle collisions (WVC) in the Czech Republic. Since 2011, the proportion of WVC has increased from 5% of all traffic crashes to almost 15%. WVC have therefore become a serious issue causing unnecessary deaths of animals, injuries to vehicle crews and property damage.
While fencing is often recommended as a highly effective WVC mitigation measure for motorways, there is a lack of a similar standard which could be applied to roads with less traffic. These roads should also remain permeable for wildlife. One currently widely applied WVC mitigation measure in Central Europe is the installation of odor repellents. The odor repellent consists of foam containing an unpleasant scent attached to a pole. There have been contradictory results, however, regarding the effectiveness of the repellents in reducing the frequency of WVC. We consequently employed an extensive experimental study focusing on ungulates to clarify this issue.
Prior to the extensive study, a pilot study was carried out in 2014 – 2016. It concluded that the installation of odor repellents reduces the frequency of WVC by at least 26% and at most 43%. These results were based, however, on only a limited sample size of found cadavers. A power analysis reached the conclusion as to how many records were needed to reach a certain probability level in detecting the effect of odor repellents. It thereby guided us on how much effort needed to be invested in the data gathering.
Benefiting from the experience achieved during the pilot study and based on the power analysis, an extensive study was planned for the period 2021 – 2022. It was designed as a before-after control-impact study using Cochran-Mantel-Haenszel method, Odds ratio and Bayesian inference to evaluate the effectiveness of the WVC mitigation measure. Our goal was to determine the effect of an at least 50% reduction of WVC, if such an effect exists, with a statistical power of at least 0.8. According to the power analysis, it turned out that at least 133 km of the road network has to be monitored. In total, 135 sections were selected and randomly split into an ‘impact’ part and a ‘control’ part. Their overall length reached 150 km.
The first phase of the study was conducted in 2021. Selected road sections were monitored by 56 field workers once per week for 7 weeks during spring or another 7 weeks during autumn. No odor repellent was applied in this phase. In the second phase, which began in March 2022, odor repellents were installed along the ‘impact’ road sections while maintaining the ‘control’ sections without any mitigation measures. The last week of the monitoring is planned for the first week of December 2022.
Our contribution focuses on the methodological issues experienced during the planning phase and when carrying out this experimental study, as well as on presenting the final results, which were unknown before the deadline for the abstract submission.