Dynamic growth and yield modeling with climate: A model for plantation Douglas-fir in the Pacific Northwest.

摘要

A pair of nested, dynamic, stand level growth and yield models with an annual base time step were developed to examine the potential effects of climate, represented by the Palmer Drought Severity Index (PDSI), on the growth of untreated plantation Douglas-fir (Pseudotsuga menziesii) stands in the Pacific Northwest. The nested models consist of a system of three autonomous, nonlinear, ordinary differential equations represented in a canonical mathematical formulation called an S-system. The models project stand density, quadratic mean diameter, and top height. The models assume only that size-density relationships exist coupling stand density and QMD, that size-size relationships exist coupling QMD and top height, and that climate modifies mortality and average growth rates.; A two step procedure was used to develop the nested models. First, a base model without climate effects was defined and calibrated. Second, annual average PDSI values were converted into a multiplicative modifier used to adjust stand mortality and average growth rates. Parameters for both models were estimated using a least squares criterion, keeping the base model parameters fixed when calibrating the climate model to isolate the climate effects.; Data from 167 untreated Douglas-fir stands containing 270 research plots, and spanning the region from southern Oregon to southern British Columbia, west of the Cascade Mountains were used to calibrate the models. The PDSI based climate modifier was then used to calibrate the extended model. The Douglas-fir measurement data spanned the growing years 1969 through 1997, which included at least one drought, with initial stand ages ranging from 4 to 113 years and measurement histories ranging from 2 to 26 years.; The base model performed well and was generally unbiased, producing stand trajectories that were indistinguishable from the data. The extended model and the base model were also indistinguishable, indicating that with these models and this data, a direct climate effect was not detectable. Several factors contributing to this result were identified, the most important being that average climate effects on mortality and growth rates are already incorporated into the model through the measurement data and the least squares parameter estimation process.