A chain of models from tree growth to properties of boards for Pinus nigra ssp. laricio Arn.: simulation using CAPSIS (cINRA) and WinEpifn (cINRA)

Both forests managers and wood industry need management tools integrating tree growth and wood quality information in order to predict the influence of site quality and silvicultural regime on individual tree size, stand yield, stand structure and wood quality. Our research project consisted in connection of models that range from growth to wood quality. This chain of models incorporates an "individual-tree, distance-independent" growth model, a branching model , a stem profile equation, log grading rules and board properties relations. This chain of models was embedded in CAPSIS software (Computer-Aided Projection of Strategies In Silviculture) which allows simulation of the growth of pure even-aged stands under different conditions (site quality, initial stand density and thinning treatments) and in WinEpifn which permits the simulation of wood properties (wood density, stiffness, heartwood content, etc.), of the grading of logs, and the sawing and the grading of boards. We present some simulations performed using this decision-making tool. Two types of simulations are carried out : "Type 1" : a real stand which was measured several times. We compare the simulated growth and the real data, on 8-year period ; "Type 2" : virtual stands on which two types of silvicultural regimes were conducted. Type 1 simulation shows a way that could be used to validate the growth model. Type 2 simulation allows to compare the effects of different silvicultural regimes on stand response and on tree characteristics. The growth model and the branching model used to perform the simulation are presented and discussed in order to emphasise the decisive importance of the data used for calibration and the effect of the model structure on the description of variability. For each tree, the stem profile and the branch characteristics can be estimated. For each plot and thinning regime, we compare the grading of logs, the board property distributions, and the volume of boards per quality class. These models and software can help to decide the most interesting approach to manage stands of #Pinus nigra# ssp. #laricio# in the centre of France taking into account the effects of site characteristics and of different silvicultural factors. A rational choice of the best silviculture can be set up considering the simulation results in terms of stand characteristics, individual characteristics of trees or/and wood quality of logs or boards