In the tropics, the conservation of tree species is closely linked to that of animals, as a large proportion of trees are zoochoric and therefore rely on the movement of animals to disperse their seeds in order to increase the chances of the survival of progeny and to allow migration in the face of climate change. Research into the prediction of animal movements is therefore critical but has so far focused only on particular features of collective behavior. In contrast, we included the concepts of cohesion maintenance, feeding area search and transient leadership in a single model, CoFee-L, and tested it to simulate the movement of a wild-ranging troop of primates (Macaca leonina). We analyzed and compared observations and simulations with a statistical physics tool (mean squared displacement) and with histograms and χ2 (for the step length and turning angle distributions). CoFee-L allowed us to simulate the physical properties of the troop's center of mass trajectory as well as the step length and angle distributions of the field data. The parametrization of CoFee-L was rather straightforward, as it was sufficient to fix a set of parameters easily observable in the field and then to adjust the values of four parameters that have biological meaning.
Keywords: collective movement; individual-based model; mean squared displacement; mechanistic modeling.