Ecological traits explain wild felid responses to human-modified landscapes in Brazil: an open-data approach for conservation

Understanding how wild felids respond to human-modified landscapes is critical for designing effective conservation strategies, yet comparative assessments across species remain scarce in tropical regions. Here, we assess the habitat selection and road sensitivity of nine wild felid species in Brazil using an integrative and scalable framework based entirely on open-access data. We compiled over 14,000 cleaned occurrence records and modelled habitat associations using generalized linear models across three spatial scales (5, 15, and 30 km). Species were grouped into three ecological categories: generalists, flexible specialists, and strict specialists, based on their ecological plasticity. Our results reveal marked interspecific variation in habitat preferences and responses to roads. While generalists like Puma concolor and Herpailurus yagouaroundi showed higher tolerance to anthropogenic environments, others like Leopardus geoffroyi exhibited more restricted patterns. Flexible specialists like Panthera onca and Leopardus pardalis avoided roads but used heterogeneous natural landscapes. Strict forest specialists (L. wiedii, L. guttulus, L. tigrinus) were highly dependent on forest cover and generally avoided open habitats. Notably, road effects were species-specific and not always aligned with ecological classification. Our findings emphasise that functional classification can aid in anticipating species vulnerabilities, but conservation actions must be tailored to species-level responses. This study highlights the potential of open biodiversity data to inform large-scale conservation planning for Neotropical carnivores, especially in data-poor regions. DOI https://doi.org/10.32942/X25P9G Subjects Ecology and Evolutionary Biology, Life Sciences

Habitat selection, Roads, Leopardus, Herpailurus, Panthera, Puma Dates Published: 2025-05-12 23:02 Last Updated: 2025-05-16 16:12 Older Versions License CC-By Attribution-NonCommercial-NoDerivatives 4.0 International Additional Metadata Language: English