Integrating drone imaging into behavioral monitoring of feedlot cattle

Continuous behavioral monitoring in feedlot systems is essential for improving animal welfare and production efficiency; however, traditional direct observation is labor-intensive and subject to observer-related limitations. This study aimed to validate the use of unmanned aerial vehicles (UAVs) to assess behavioral responses of beef cattle in a commercial feedlot. Data were collected over 21 non-consecutive days in 34 pens housing up to 150 animals each. A DJI Mini 2 drone recorded aerial images at altitudes between 10 and 20 m during the hottest hours of the day (13:00–15:00 h), while simultaneous direct visual observations were performed as the reference method. Behaviors were classified according to a predefined ethogram. Agreement between methods was evaluated using the Intraclass Correlation Coefficient (ICC), based on the number of animals exhibiting each behavior per 10-minute observation session recorded by each method. Excellent agreement was observed for eating behavior (ICC = 0.90), while standing (ICC = 0.86), lying (ICC = 0.80), and dominance/mounting behavior (ICC = 0.79) showed good reproducibility. In contrast, drinking (ICC = 0.48), ruminating (ICC = 0.41), and walking (ICC = 0.37) presented poor agreement. Higher reliability was associated with behaviors of longer duration and clearer postural definition, whereas dynamic and short-duration behaviors were more affected by visual and methodological constraints. Overall, UAV-based monitoring proved to be a reliable and non-invasive tool for assessing key behavioral patterns in feedlot cattle, particularly feeding activity, supporting its application in precision livestock farming systems.