I recently published a paper with Paul McDonald and Graham Hall at University of New England in Australia where we evaluated the potential of UAVs (drones) for large scale waterfowl surveys. Most large scale waterfowl surveys are currently done using fixed wing aircraft with one or two trained observers looking out a window and counting/identifying birds as they go. We thought we could harness some of the power of UAVs to provide detailed photographic surveys of targeted wetlands without disturbing birds and providing an archive of high resolution photos that can be independently verified and perhaps even used for automated species detection. There is a lot of excitement out there about the use of drones for ecological applications but not a lot of rigorous testing of different types of drones in wild situations to see if they cause disturbance to target species. After a lot of trial and error and working with different shapes and sizes of drones we found that if you are properly prepared and think carefully about take-off and landing sites, approach angles and flight altitudes, there should be minimal disturbance to wild flocks of birds. Birds only took flight if the UAV was launched directly towards them or dropped altitude or banked suddenly, particularly if the wing profile of the UAV mimicked that of an avian predator. The cameras we used allowed us to unambiguously identify even small sized birds like grebes. In some shots we could make out the eye colour of individual birds. To read the full paper click here.
