How animals adapt their movements to take advantage of behaviorally-relevant time intervals is not well understood, especially in the supra-second timescale. It has been proposed that motor timing depends on the emergence of self-sustained dynamics across ensembles of neurons. Alternatively, evidence from operant conditioning suggests that animals can develop motor routines to adapt their behavior to fixed temporal constraints. But it is unclear whether animals can accurately time their behavior without the help of motor routines. To address this issue, we used a task in which rats, freely moving on a motorized treadmill, could obtain a reward if they approached it after a fixed interval. Most animals took advantage of the treadmill length and its moving direction to develop, by trial-and-error, a unique motor routine whose execution resulted in the precise timing of their reward approaches. Noticeably, when proficient animals occasionally failed to follow this routine, the timing of their reward approaches was systematically poor. In a second step, we trained naive animals in modified versions of the task specifically designed to prevent the development of this motor strategy. Compared to rats trained in the first protocol, these animals never reached a comparable level of timing accuracy. We conclude that motor timing critically depends on the ability of animals to develop motor routines adapted to the structure of their environment. Our work also suggests that self-sustained neuronal activity alone may not be sufficient to support motor timing, at least in the supra-second timescale.
