Presented by Jacob Bushaw - Email: firstname.lastname@example.orgDuck pair counts, nest monitoring, and brood surveys are important for estimating abundance and productivity on the breeding grounds. These time-consuming surveys require substantial investments in personnel and are still prone to systematic non-detections. Recent advances in Unmanned Aerial Vehicles (UAVs) have made it possible to conduct aerial surveys for a variety of wildlife; here, we evaluated their utility for conducting breeding duck surveys in southern Manitoba. We conducted pair counts with a UAV equipped with visual and thermal cameras and located 237 breeding pairs. While UAV surveys were substantially faster than ground surveys, they located fewer pairs then the ground crew because it was difficult to distinguish social groupings from the air. We searched for overwater nests with the UAV on 48 quarter-sections: 24 were also searched by ground technicians and 24 were searched and monitored only with the UAV to test for ground observer effects on nest success. We located 40 nests on drone-only quarters and nest success was 24.3%, compared to 28.7% for nests monitored on other sites by ground technicians. We located 125 nests on sites where crews overlapped: 74 nests were located only by the ground technicians, 23 nests were located by the UAV and missed by ground technicians, and 28 nests were located by both crews. Thick vegetative cover impaired the ability of the UAV to locate nests, and it was difficult to find nests not being incubated. During the two rounds of brood surveys the UAV located 669 broods with a detection rate of ~56%, compared to 334 broods located by the ground technicians with a detection rate of ~27%. We conclude that UAVs are ineffective for pair surveys, show some promise for locating overwater nests, but are twice as effective as traditional brood surveys.