Roads are currently one of the main attributes of human-dominated ecosystems. Aimed to improve the socio-economic conditions of local communities, the global road network is exponentially developing associated to the increase of gross domestic product in several world regions. Roads have negative impacts on wildlife, and these impacts are studied by Road Ecology. However, studies to date are mainly focused on individual-level approaches, even though population-level studies have greater power to understand road impacts, as they focus on the persistence of whole populations. Here, we review the literature on road-kill and fragmentation studies, and their corresponding mitigation measures, emphasizing those employing population-oriented approaches. Road-kill rates are especially high when species occurrence and behaviour matches high traffic flow and speed. Those species that do not avoid roads undergo higher mortality rates. Species abundance, foraging behaviour or being ectotherms, make species more susceptible of being road-killed. On the contrary, species that avoid roads are more prone to undergo connectivity reduction. This avoidance can be due to the road itself, to vehicle emissions or to the vehicles themselves. It is expected that those traits influencing the impact of general habitat fragmentation, like low reproductive rates, long generation time or large area requirements, are also important for road-mediated fragmentation. Fencing road sections is the most effective mitigation measure to reduce road-kills. Unfortunately, fencing also reduces habitat connectivity which developers try to minimize with the construction of wildlife passages. Some methodologies which explore the effectiveness of mitigations, like carcass counts or passage use, do not provide information on their effectiveness for population persistence, which should be a priority of research. Rather than limiting data collection to detect road mortality, reduced connectivity or passage use, road ecologists should strive to quantify whether ecosystem functionality is being altered by roads, and whether it has been restored (if necessary) after mitigation actions. For these purposes, a priority should be to explore the demographic impact of roads on population persistence probability, and how this increases after mitigation. To this end, genetics are useful to identify population substructuring due to road presence, to know effective population sizes, breeding rates or to identify individuals. Population modelling using demographic approaches can be employed to explore how populations respond to road mortality, to select the least impacting construction alternative, as well as to identify the sections where mitigation measures maximizes the functional connectivity. Finally, the huge development of road networks in environmentally vulnerable countries will benefit from the input of ecologists during the planning process, from the integration of roads within Strategic Environmental and Social Assessments, and from the inclusion of targets explicitly focused on linear infrastructures on the Strategic Plan for Biodiversity by the United Nations.
