Modelling the spatial structure of heterogeneous stands. Example of Oak-Scots pine mixed stands of the Orleans forest (France)

To model the growth of heterogeneous (uneven-aged or mixed) stands, the use of individual based models seems more appropriate, because of the individual variability within heterogeneous stands. However, forest managers cannot easily use this kind of models, especially because they require knowing the localisation of each tree in order to run simulations. However, it is possible to use our knowledge on stand level characteristics for a forest stand as rules to build a virtual stand, which will be close to the real stand, and which could be used as initial state for an individual based model. This approach corresponds to the development of a model of spatial structure. Our aim is to explain the principle of the development of such a model, and to illustrate it on the case of oak- Scots pine mixed stands. First step consists on characterizing the spatial structure of the studied stand. In our example, we built a typology of spatial structure of these stands; we will present the main spatial types identified. Second step specifies how to use adapted statistical tools - point processes - to build a model of spatial structure from the identified types. Finally, we will show how such a model of structure would help simulating realistic initial states, which would facilitate the use of individual based growth model.