When road-kill hotspots do not indicate the best sites for road-kill mitigation.
The effectiveness of measures installed to mitigate wildlife road-kill depends on their placement along the road. Road-kill hotspots are frequently used to identify priority locations for mitigation measures. However, in situations where previous road mortality has reduced population size, road-kill hotspots may not indicate the best sites for mitigation. The purpose of this study was to identify circumstances in which road-kill hotspots are not appropriate indicators for the selection of the best road-kill mitigation sites. We predicted that: (i) road-kill hotspots can move in time from high-traffic road segments to low-traffic segments, due to population depression near the high-traffic segment caused by road mortality; (ii) this shift will occur earlier for more mobile species because they should interact more often with the road; (iii) this shift can occur even if the low-traffic segment runs through lower quality habitat than the high-traffic segment. To test these predictions, we simulated population size and road-kill over time for two populations, one exposed to a road segment with high traffic and the other to a road segment with low traffic. Our simulation results supported Predictions 1 and 3, while Prediction 2 was not supported. Synthesis and applications. Our results indicate that, for new roads, road-kill hotspots can be useful to indicate appropriate sites for mitigation. On older roads, road-kill hotspots may not indicate the best sites for road mitigation due to population depression caused by road mortality. Direct measures of the road impact on the population, such as per capita mortality, are better indicators of appropriate mitigation sites than road-kill hotspots.