Link to thesis
Abstract in English Since mortality rates observed in new cork oak plantations are relevant for future productivity, it is important to be able to find out the locations where cork oak finds best conditions for development. As these conditions are related to edaphic, soil and climate local parameters, this work aims at modeling the proportion of dead trees in young plantations of cork oak, using a set of permanent plots established in different regions of Portugal (Coruche, Évora, Ponte de Sor and Castelo Branco). Three logistic regression models that allow the prediction of the proportion of dead trees (%) as a function of soil and/or climate variables were fitted, their difference lying in the need of opening a soil profile and the capacity to determine the presence of a waterproof layer were adjusted. The resulting parameters showed a reduction of dead tree proportion in stands when higher the site index, higher soil depth and in the presence of medium soil textures. The increase of the proportion of dead trees with the increase of medium annual temperature and in the presence of a waterproof layer was also demonstrated. The adjusted models will be included in the SUBER growth and yield model.