Modeling percent stocking changes for lodgepole pine stands in Alberta

Abstract

A percent stocking change model was developed for lodgepole pine ( Pinus contorta Dougl. ex Loud. var. latifolia Engelm.) in Alberta based on spatially mapped permanent sample plot data. Percent stocking was defined as the percentage of 10 m 2 subplots occupied by at least one tree with a minimum height of 1.3 m. The difference equation technique was employed to fit the model. Three model forms were examined and the logistic function was chosen as the final model. Site index was found to be a significant predictor and incorporated into the model. Analyses revealed that the model had correlated, but homoskedastic errors and the correlated errors were modeled by spherical covariance structure using NLINMIX macro in SAS. A percent stocking index, defined as the percent stocking at 50 years total age, was introduced and derived from the developed model. The percent stocking model had both forward and backward projection capabilities. It was demonstrated, both on model fitting and validation data, that the model adequately portrayed the percent stocking dynamics of lodgepole pine stands in Alberta. The model also provided an important basis for creating linkages between reforestation survey results and future yield, which is crucial for sustainable forest management.