Unsupervised Ethograms of a Vulnerable Bird Species: The Red-Footed Falcon in Northern Italy

Abstract

Behavioral and movement ecology have quickly advanced as a result of the development of biotelemetry devices and analytical techniques. Global positioning system (GPS) transmitters assist scientists in gathering location and movement data at detailed spatial and temporal resolutions. Machine-learning methods can then be applied to GPS data to provide insights into the ecological mechanisms of animal behavior and movements. By means of accurate GPS data-loggers, in 2019, 2020, and 2021, we tracked 8 red-footed falcons at the two largest colonies in Italy. We collected 13,484 GPS points and used recently introduced machine-learning methodology Unsupervised Animal Behaviour Examiner (UABE) to deduce the regular, nested, and hourly ethograms of the tracked individuals. We found clear and significant patterns of the red-footed falcons’ behaviors on monthly, daily, and hourly bases. Our study is a step forward in advancing the knowledge of this threatened species, and provides a baseline assessment of the current behavioral patterns of this red-footed falcon population, with which results of future studies can be compared to detect potential behavioral changes that act as early warnings of increased human disturbance.