Predictive mapping for forest sites - A new tool for forest managers

摘要

Spatial data layers and a phytoecological database were used to achieve a homogeneous mapping of the forest site types over a broad area (Vosges Mountains, Northeast France). Three main ecological factors defining forest types were firstly identified through the study of forest site classifications: soil nutrient content, soil moisture and bioclimate. Synthetic variables with these three factors were computed through the analysis of floristic and ecological data available in the area investigated. These three synthetic variables were then modelled with ecological layers using a Geographic Information System to obtain three elementary maps of soil nutrient content, soil moisture and bioclimate. A GIS combination of these maps provided the numerical map of ecological units homogeneous on the scale of the Vosges Mountains. The integration of temperature and water balance in the bioclimatic model will enable climate changes to be taken into account and so update future maps of ecological units. The validity of the predictive map of forest sites was checked and it was compared to mapping carried out in the field. The distribution of fir and sycamore in the site types was assessed in order to verify the link between the forest site types predicted by this new site classification method and the forest site potential.