Unoccupied aerial vehicles as a tool to map lizard operative temperature in tropical environments
Emma A. Higgins,Doreen S. Boyd,Tom W. Brown,Sarah C. Owen,Geertje M. F. van der Heijden,Adam C. Algar
DOI: https://doi.org/10.1002/rse2.393
IF: 5.7874
2024-04-27
Remote Sensing in Ecology and Conservation
Abstract:This study outlines a new workflow, using UAVs, to map lizard operative temperatures across landscapes at organism‐relevant spatial resolutions, which until now was not possible. This workflow allows us to measure thermal suitability and heterogeneity across continuous landscapes, which is crucial to inform conservation actions for ectothermic species. Here, we use random forest models to couple UAV‐derived canopy measures (texture and greenness) from RGB imagery, with sub‐canopy thermal 3D replicas, to predict lizard operative temperature at solar noon. This workflow was found to work well across forest, mangrove and urban environments and allows us to measure thermal suitability across whole areas, rather than solely point‐based measures from standard thermal 3D replica methods. This work provides a simple workflow to measure a directly ecologically relevant measure at relevant spatial scales, across continuous surfaces, and allows us to consider thermal habitat suitability, energetics and climate change risk for ectotherms. To understand how ectotherms will respond to warming temperatures, we require information on thermal habitat quality at spatial resolutions and extents relevant to the organism. Measuring thermal habitat quality is either limited to small spatial extents, such as with ground‐based 3D operative temperature (Te) replicas, representing the temperature of the animal at equilibrium with its environment, or is based on microclimate derived from physical models that use land cover variables and downscale coarse climate data. We draw on aspects of both these approaches and test the ability of unoccupied aerial vehicle (UAV) data (optical RGB) to predict fine‐scale heterogeneity in sub‐canopy lizard (Anolis bicaorum) Te in tropical forest using random forest models. Anolis bicaorum is an endemic, critically endangered, species, facing significant threats of habitat loss and degradation, and work was conducted as part of a larger project. Our findings indicate that a model incorporating solely air temperature, measured at the centre of the 20 × 20 m plot, and ground‐based leaf area index (LAI) measurements, measured at directly above the 3D replica, predicted Te well. However, a model with air temperature and UAV‐derived canopy metrics performed slightly better with the added advantage of enabling the mapping of Te with continuous spatial extent at high spatial resolutions, across the whole of the UAV orthomosaic, allowing us to capture and map Te across the whole of the survey plot, rather than purely at 3D replica locations. Our work provides a feasible workflow to map sub‐canopy lizard Te in tropical environments at spatial scales relevant to the organism, and across continuous areas. This can be applied to other species and can represent species within the same community that have evolved a similar thermal niche. Such methods will be imperative in risk modelling of such species to anthropogenic land cover and climate change.
ecology,remote sensing