Análise de dados de atropelamentos de mamíferos em um trecho da BR-471, Rio Grande do Sul

• Ano de defesa: 2022
• Autor(a) principal: Cassenote, Maite De Dineque
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/26339/0013000016cr1
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Engenharia Ambiental; UFSM; Programa de Pós-Graduação em Engenharia Ambiental; Centro de Tecnologia
• Palavras-chave em Português: Ecologia de estradas; Acidentes de tráfego; Mortalidade rodoviária; Fatores ambientais; Modelos de regressão; Vida selvagem; Ecology of roads; Traffic accidents; Road fatalities; Environmental factors; Regression models; Wildlife; CNPQ::ENGENHARIAS
• Link de acesso: http://repositorio.ufsm.br/handle/1/27349

Faced with great economic demands, road infrastructure performs an important economic service. Its environmental impacts affect processes and natural resources, among which is the trampling of wild animals. The collisions of vehicles with wildlife develop a series of conflicts that go far beyond the reduction of fauna, such as road safety problems arising from inadequate management. The fundamental stage for the development of mitigation solutions is the monitoring and recording of animal collisions. In Brazil, wildlife monitoring data available in free format, updated, faultless and consistent are rare. The initial structure of a consistent database demands time, trained personnel to identify species and, mainly, financial resources. This study used an open-access database, made available by researchers, which contains records of pedestrian collisions on Brazilian roads. For this study, we considered the BR-471 highway, known for conducting traffic accidents involving species in the southern region of Brazil. The aim of the study was to analyse the records of mammalian road kills, by means of generalised additive models for position, scale and shape (GAMLSS), non-parametric tests, descriptive analyses and the use of a geographical information system. Furthermore, with the location of the road kills monitoring it was possible to optimize the existing database with data on population abundance of living species, land cover, precipitation and temperature, among other factors that could represent specific conditions of road kills. Therefore, we conclude that animal counts, temperature, seasonality, area surrounding the roadway, among other environmental factors can be considered in the discussion on road kills of mammalian species. Therefore, the data analysis provides a series of discussions regarding the species that are run over, pointing out specific conditions of road kills in relation to environmental variables, i.e. factors that can be incorporated into road planning and management.