In this paper, we present a contribution to time and space management in reactive simulations, especially for crowd simulation in an urban or closed environment. We are not interested in mere individual scheduling, but rather in the link between global rhythm issues and individual behavior. Our aim is to simulate real moving behaviours in constrained environments, taking into account the link between the macroscopic predictions of the simulation and experimental data. The simulation must deal with collision problems, rhythms variations or multiple moving models. We argue here that a synthesis between temporal and spatial features is needed. This work relies upon our MISC model which allows to simulate 2D-moves with different rhythms according to population classes, and gives a statistical control on the life cycle of the individuals.