High-resolution drone imagery reveals drivers of fine-scale giant otter habitat selection in the land-water interface
Valladares NA., Pardo AA., Chiaverini L., Groenendijk J., Harrington LA., Macdonald DW., Swaisgood RR., Barocas A.
Neotropical freshwater habitats are particularly sensitive to degradation by human activity. Piscivorous semi-aquatic freshwater megafauna inhabit both the terrestrial and aquatic mediums and thus may be good indicators of wetland habitat quality. However, the drivers of their space use at the terrestrial and aquatic landscape levels are not well understood. We studied the spatial behavior and habitat use of giant otters in Madre de Dios, Peru, inhabiting areas with variable levels of protection. We combined unmanned aerial vehicle (UAV) and satellite images to develop different terrestrial and water-associated land cover variables. We tested the influence of these predictors on giant otter habitat use at multiple spatial scales, comparing used and available locations. Giant otters favored bank areas with dense forest canopy cover. In the aquatic medium, giant otters showed positive selection for open water and fallen logs and avoided floating vegetation. These findings may be explained by preference for optimal fish habitat to maximize foraging yield and bank areas that provide more cover from predators and higher quality denning locations. Variables developed from UAV images outperformed satellite-derived variables. Despite recent signs of deforestation in lake banks in unprotected areas, spatial model predictions indicated that unprotected oxbow lakes did not differ in their habitat suitability from protected freshwater habitats. Management implications of our findings include identification of factors driving habitat suitability to guide policy and decisions regarding protection or restoration of oxbow lake ecosystems to support giant otter populations. In addition, we demonstrate that UAVs have value in complementing satellite-derived images and providing a cost-effective methodology to assess habitat quality for semi-aquatic species at the land-water interface.