Abstract
Understanding how birds adjust their flight in response to biomechanical characteristics and environmental conditions can be useful for interpreting homing behavior. This study investigates homing pigeons' (Columba livia) flight behavior using multi-sensor biologgers, integrating GPS, tri-axial accelerometer, pressure, and temperature sensors. Flight biomechanics were assessed by extracting: wingbeat frequency from the Short-Time Fourier Transform of the total acceleration signal and peak-to-peak acceleration from the dorso-ventral component. Landscape characteristics were provided by classifying land cover along the route using a geographic atlas and by computing flight altitude above ground level through the combination of pressure-derived altitude and a digital elevation model. The results reveal a progressive decrease in wingbeat frequency along the homing route, showing a linear relationship with traveled distance. To assess whether this pattern can be interpreted in terms of flight regulation, flight altitude was modeled as a function of biomechanical and environmental variables using a linear mixed-effect approach. The analysis indicates that flight altitude is significantly affected by wingbeat frequency as well as by temperature, ground speed, and land cover, with wingbeat frequency and temperature showing the strongest negative association.