Spatial and temporal influences of road duplication on wildlife road kill using habitat suitability models
MetadataMostrar registro completo
Urban growth and population growth led to the construction of a gigantic road network around the world. This network is responsible for several impacts on fauna, flora and the environment, such as road kill, isolation of populations, facilitating the establishment of invasive species, river siltation, among others. However, although road ecology has advanced recently, there are still many gaps on how roads affect fauna, as little is known about how effects of changing the structure of highways can modify animal-vehicle collisions. This study aims to evaluate some of the impacts of roads on species of medium and large mammals in fragmented and naturally heterogeneous landscapes. Using a functional group approach based on animal sensitivity to disturbance and displacement capacity, I set out to answer three questions, one in each chapter: 1) the contribution of various landscape indices to predict wildlife road kill; 2) highway duplication and the implementation of wildlife crossing structures alter animal road kill; 3) duplication of roads change the way fauna road kill is correlated with the landscape metrics. To answer the first and third questions, we have developed innovative methods combining road kill data with a multi-scale approach with landscape metrics involving quantity and distance of various landscape elements, such as natural vegetation, cerrado, water, forestry and sugar cane. This method proposed was derived from habitat suitability model, and proved very promising for estimating the probability of animal road kill. Each functional group of species responded differently to landscape elements. Distance and amount of vegetation has been more important to estimate road kill probability of more sensitive mammals, but the amount of sugar cane also contributed to these results. The proposed method is highly replicable and can be easily applied in other regions with other taxa. The second question was addressed in an analytical way, with a conventional hypothesis testing approach. We found that, in general, there was no significant difference between road kill before and after road duplication. However, when considering the functional groups, and even species, some changes were significant for both increasing and reducing road kill. We also found that the proximity of wildlife crossing structures to road kill records did not reduce the frequency of animal-vehicle collision, indicating that such mitigation measures may not have been appropriate to reduce animal road mortality. Finally, in the third chapter we have proposed a new approach to estimate the changes in animal road kill probability before and after duplication of highways. In this chapter we recorded an increase in the probability of road kill after duplication for generalist species with high mobility. The methods proposed here are easy to implement in several actions related to roads, both for seeking their structural improvement and for making them more sustainable for biodiversity.