Thermal drones for wildlife research in tropical forests: A review of best practices, challenges, and opportunities

Tropical forests support exceptional biodiversity yet remain underrepresented in long-term wildlife research and monitoring programs due to logistical, financial, and methodological constraints. Conventional ground-based survey methods are often ineffective in these environments, particularly for rare, cryptic, or arboreal species. Drone-mounted thermal cameras (thermal drones) offer a promising alternative by detecting endothermic animals from above based on their thermal signatures, with advantages for surveying inaccessible or densely forested habitats. This review synthesises the application of thermal drones for research on tropical forest fauna and provides a practical framework to support conservation practitioners in their deployment. Drawing on 38 studies published between 2018 and 2025, we examine how biological traits, environmental conditions, and technical parameters influence detection outcomes and evaluate whether thermal drones have potential to improve upon conventional methods. While most studies remain at the pilot stage, with limited progression to population inference, the available evidence indicates that thermal drones can outperform ground-based methods for detecting arboreal mammals in tropical forests when protocols are tailored to species behaviour and conditions that maximise thermal contrast. Based on the evidence, we summarise operational challenges and methodological limitations and highlight key constraints on inference, including variable detectability and species misclassification. We distil findings from the literature to provide best practice guidance for survey planning, flight configuration, and detection validation, supported by a decision tree to guide protocol design. Integration with long-term monitoring programs, improved error quantification, and broader taxonomic application will be essential to realise the potential of thermal drones for tropical biodiversity research.