Habitat suitability models (HSMs) are essential for managing threatened species, but often overlook behaviour-specific habitat needs due to the difficulty of obtaining spatially explicit behavioural data from wild animals. This can lead to misidentification of suitable habitats and misinterpreting habitat selection processes. Contemporary multi-sensor biologgers, which provide unprecedented behavioural insight, are under-utilised in this context.
Biologging data from flatback turtles Natator depressus (n =42) at a macrotidal embayment in Western Australia were used to identify and characterise suitable habitat for key in-water behaviours. Foraging and resting locations derived from high-resolution motion sensor data (accelerometer, magnetometer) coupled with animal-borne video and GPS data, were combined with 10 environmental features in a series of random forest HSMs (AUC >0.84).
Bathymetry, distance from the coast, and currents were key features determining habitat suitability, with different influences on foraging and resting behaviours. Suitable habitats for both behaviours were generally shallow (10–15m) and nearshore (5–10km from the coast), but areas within 5km of the coast were more suitable for foraging than resting. Habitat suitability and overlap with spatially zoned areas varied with water level, reflecting tidal influences and highlighting dynamic management opportunities. Our generalisable approach enhances the utility of biologging and HSMs for decision makers.