Fitting and calibrating a three-level mixed effects cork growth model

Cork value is determined mainly by its thickness at the moment of debarking but also affected presence any discontinuities related to porosity characteristics and/or anomalies in cork plank. quality closely final price raw material, so inventories conducted prior are aimed identifying and therefore potential price. Besides information from these inventories, forest managers need a reliable prediction help them determine most profitable harvest. In this work, three-level nonlinear mixed effects growth model was developed for Spanish oak forests using 1027 samples taken nine regions provenance into which Spain divided identification reproductive material. The fitting step revealed that region stand level explained much less variability than tree mixed-effects models. Therefore, subject–specific predictions were obtained with joint use region-, stand-and tree-level random effects. compared three alternative options when no additional measurements available. precision calibrated stand-, 0.1 mm taking one measurement time during rotation period. more precise has wider range application previous model, being suitable predicting accumulated 90% distribution area Spain. However, it should only be applied make within given period not simulate over successive rotations. future, would necessary develop an annual includes site, variables capable providing long-term under different climatic management scenarios. study could provide basis such because used identify main through correlation analysis.