Background
In the current context of climate change and intensive anthropization of natural environments, the preservation of ecosystems is a matter of great importance. In the tropics, the conservation of tree species is closely linked to that of animals. A large proportion of tree species is zoochore and needs animals, mainly vertebrates, to disperse their seeds in order to increase their chances of survival. Being able to predict the movement of animals is therefore an important skill. Lots of research exists on the subject, but it has always focused on particular characteristics of collective behavior. In contrast to previous studies, we included the concepts of Cohesion maintenance, Feeding area search and transient Leadership in a single model, CoFee-L.
Methods
We tested CoFee-L to simulate the movement of a wild-ranging troop of primates belonging to the species Macaca leonina in a nature reserve of Thailand. We analyzed this methodology by statistical physics tools (mean squared displacement) and we characterized the simulated trajectories by histograms and χ2 for the step length and turning angle distributions sampled at the same frequency as the collection of field observations.
Results
CoFee-L allowed us to reproduce the physical properties of the troop's centre of mass trajectory, as well as the step length and angle distributions of the field data. Moreover, CoFee-L is able to produce trajectories looking like correlated random walks or levy walks depending of the parametrisation.
Conclusions
We created a model, CoFee-L, which allowed us to simulate the movement of a group of vertebrates taking into account not only the movement and individual characteristics of each member, but this also independently of the group size. The parameterisation of CoFee-L was rather simple, as it was sufficient to fix a set of parameters easily observable in the field (number of individuals, size of the study site, duration of monitoring, duration of foraging, distance between individuals and maximum spread of the group) and then to adjust the values of 4 parameters which have biological meaning (personal information, exploration of the study site, speed of movement and tendency to do a levy walk).