Atropellos de mamíferos en la Región Este de Uruguay y su relación con atributos del paisaje

Abstract

Roads constitute a threat to biodiversity, causing negative impacts such as fragmentation of habitats and run overs. In 2015, the NGO ECOBIO Uruguay began working on the subject in order to implement mitigation measures. The objective of this work was to describe and analyze the run overs of medium and large mammals for the Eastern Region of Uruguay and their relationship with the landscape's attributes. Taking off the dynamic of run overs, this is frequently influenced by anthropic cases, traffic is the most prominent variable over this.  Machine learning algorithms (Random Forest) and QGIS and R programs were used to model the run overs of the region, that is characterized by having heterogeneous ecosystems, important biodiversity and several protected areas. 976 cases of run overs were analyzed, from 17 species that were distributed heterogeneously. The models analyzed by species had a predictive ability of 80% success, while when working with all species as a whole 60%. The route with the highest probability run overs was 'Route 9'. Although the most important variables were similar in all the models, the three that better illustrated the prediction models were: annual average daily traffic, the distance to backroads and distance to populated roads, this three illustrates better the predictions models. This work is one of the first studies at a national level which formally quantifies the phenomenon and suggests analyzing the species separately to evaluate possible differential responses that make it impossible to find general patterns.