Modelling movement trajectories to infer behavioural states and variation in the undulate skate, Raja undulata

The causes and consequences of fish movement behaviour are critical yet understudied aspects of aquatic species’ movement ecology. We investigated the behavioural states and individual variation in the movement of the undulate skate, Raja undulata, in Galicia, Spain. For this we acoustically tracked 176 individuals over a period of four years. We then calculated the trajectories using a movement model, analysed the ecological drivers of these trajectories using Hidden Markov Models (HMMs), and investigated the existence of consistent variation in three trajectories’ properties’: sinuosity, straightness and speed. HMMs successfully identified two behavioural states, resting and active, which were influenced by day-night period (more active at night) and bottom habitats (more active in sandy bottoms). We found moderate individual consistency in speed (repeatability 0.22) compared to sinuosity (0.09) and straightness (0.07), suggesting a higher potential to evolve for speed. This knowledge can be used to protect important habitats for the species, such as the transition habitat between sandy habitat and rocky reefs, to design and implement marine protected areas, and to establish fisheries regulations, such as regulating fishing at night when undulate skates are more active.