Evaluation of the Efficacy of Wildlife Warning Reflectors to Mitigate Wildlife-Vehicle Collisions on Roads
von Anke Benten
Datum der mündl. Prüfung:2018-09-07
Erschienen:2018-12-04
Betreuer:Prof. Dr. Christian Ammer
Gutachter:Prof. Dr. Torsten Hothorn
Gutachter:Prof. Dr. Niko Balkenhol
Dateien
Name:Dissertation_ABenten.pdf
Size:2.84Mb
Format:PDF
Zusammenfassung
Englisch
Summary Wildlife-vehicle collisions cause human fatalities as well as economic and ecological losses on roads worldwide. Multiple mitigation measures have been developed over the past decades, while wildlife warning reflectors for preventing animals from entering the road when vehicles approach enjoy great popularity due to their commendableness, manageability and comprehensive applicability. However, their efficacy is still in question because of contradictory study outcomes and also behavioral studies could not find any long-term reaction of animals in the presence of the reflectors that would reduce the number of collisions. The task of this thesis, within the framework of a large-scale project initiated by the Germany Insurance Association (GDV), was to objectively analyze contradictions in literature, to evaluate the influence of modern reflectors on collisions with wildlife and on wildlife behavior. In a first study, a comprehensive literature survey was carried out to evaluate disaccords in previous studies as well as other methodological differences that might explain the variation in study outcomes. The effect size of wildlife warning reflectors on the frequency of wildlife-vehicle collisions across all available data was assessed within a metaanalysis. The meta-analysis on literature data revealed that only studies applying a beforeafter design, effective study duration of < 12 months and considerably short testing site lengths of < 5 km found an effect of the reflectors. Our second and main study focuses on the efficacy of modern wildlife warning reflectors to mitigate wildlife-vehicle collisions on roads. Three different optic reflector types that are most widely spread in Germany, as well as one opto-acoustic model were tested on 151 testing sites of approximately 2 km each in a prospective, randomized non-superiority cross-over study design for 24 months. Our results show that wildlife warning reflectors did not lower the number of collisions with ungulates by a relevant amount. Finally, since few studies as well as hunters and manufacturers have reported a potential short-term effect of the reflectors on animal and driver behavior, we tested the reaction of ungulates towards oncoming vehicles and drivers to animals near the road considering a potential habituation effect in a third study. We could not find any behavioral response of ungulates or humans with reflectors present that would have lowered the risk of a collision, as the devices did not influence the reaction of animals to oncoming vehicles or motorists to wildlife near roads from the very beginning. Considering the results of our first study, applying study designs without controlling for other, confounding factors such as a before-after study design, is not appropriate for evaluating the impact of an intervention due to the lack of independence from different levels of single treatments and true replications. A potential change after the implementation of a treatment cannot simply be assigned to that impact but to other factors as well. Moreover, the constant frequency of wildlife-vehicle collisions and the invariable responses of wildlife to approaching vehicles in the presence of the reflectors in our second and third study might be caused by the visual abilities of non-human mammals as well as the reflective properties of the reflectors. It has been shown for crepuscular and nocturnal animals that they are dichromatic, i.e. they cannot perceive long-wave light, and visual adaptations to rapid increases in light intensity such as headlights of approaching is considerably slow. Additionally, the light reflected from wildlife warning reflectors is already very low at close distances near the devices and overlaid by the headlights of approaching vehicles. Under these conditions, a potential efficacy of the reflectors is questionable anyway. Based on our results, we conclude that wildlife warning reflectors are not effective for mitigating wildlife-vehicle collisions on roads.
Keywords: Wildlife Warning Reflectors; Deer mirrors; Ungulate Roadside Behavior; Animal-Vehicle Collisions; Deer-Vehicle Collisions