Search

Large herbivores possess high dispersal rates and require vast areas to roam due to their ecology. This will make them susceptible to anthropogenic threats such as vehicle-collisions. Persian onager (Equus hemionus onager), as the only representative of Artiodactyla in Iran, is not an exception. Persian onager-vehicle collision can not only be lethal for themselves but also for motorists. Given the importance of this twofold issue, an important step being taken to reduce collisions was the installation of signs that warn motorists of the high probability of onager-vehicle collisions. We developed a questionnaire to (1) assess the effectiveness of warning signs from motorists’ perspective, and (2) to identify the most important factors affecting motorist beliefs in the effectiveness of warning signs. We solicited responses to our questionnaire from motorists on a road with a high Persian onager-vehicle-collision rate in Southern Iran (Hassan Abad-Meshkan Road). To identify factors affecting motorists' beliefs in the effectiveness of warning signs we used logistic regression and for classifying motorists’ beliefs in the effectiveness of warning signs we used decision tree. Our result showed that motorists' driving speed, lack of adequate safety equipment on the road (e.g. light, police camera), using cellphone while driving, and concern about wildlife damage while driving on the road were the significant factors affecting motorists' beliefs toward the effectiveness of Persian onager warning signs. It is necessary to increase road safety equipment, install standard warning signs at the Persian onager crossing points, and study the behavior of motorists and the rate of road casualties after the mitigation methods to protect this species.

Humans are responsible for over a quarter of all wildlife mortality events across the globe. The pressure this puts on wildlife populations contributes to the decline of many at-risk species. To minimize human-caused mortality and reverse population declines in species across the world, we first need to know where these events are happening or likely to occur since managers and public agencies often have limited resources to devote to a problem. As such, our objective was to develop a modeling approach to delineate human-caused wildlife mortality hotspots in regions with limited data. We used internet search engines and national media to collect data on brown bear (Ursus arctos) mortality events in Iran from 2004 to 2019. We then developed a spatially-explicit Maximum Entropy (MaxEnt) model using anthropogenic and environmental variables to predict the probability of human-caused brown bear mortality. We were able to delineate 7000 km2 as human-caused mortality hotspots, along with the geographical locations of those hotspots. This provides information that can help identify where critical conflict mitigation efforts need to be implemented to reduce the potential for human-caused wildlife mortality. However, more targeted studies such as surveys of local people will be needed inside hotspots identified with this methodology to assess the attitudes of humans toward different wildlife species, informing the specific mitigation actions that will need to be made. Finally, we suggest that media data can be used to identify these hotspots in regions where systematic data is lacking.