Permanent grasslands host a high plant diversity, which sustains many ecosystem services. Thus, understanding how composition of the plant community responds to different management practices under given soil and climatic conditions is crucial for making best use of grasslands. modeling approaches may be used to explain the manifold interactions involved to sustain this diversity. We developed the dynamic, process-based ecological model DynaGraM to simulate the seasonal aboveground vegetation dynamics of semi-natural grasslands. The model allows specifying plant community by any number of species. The predictive power of the model was assessed by simulating the dynamics of a virtual mountain grassland in response to four typical management scenarios under constant climatic conditions over several decades. In our experiments, we modelled an assemblage of seven species representing contrasted plant functional types. We compared model outputs to compositions inferred from floristic records conducted in the French Jura Mountains. Irrespective of initial conditions, the simulated community converged to four distinct compositions that primarily reflected management. Overall, the results matched the functional composition inferred for each of the scenarios from the botanical records. Convergence in functional composition was reached in less than 15 years under grazing scenarios, but not less than 50 years under mowing scenarios. At quasi-equilibrium, the highest vegetation diversity was obtained for extensive grazing and the lowest for extensive mowing. Overall, this study introduces a novel and relatively simple approach to model competition and adaptation processes in plant community dynamics, thus providing a response to the key challenge of modeling multi-species grasslands.
