Height-diameter models with random coefficients and site variables for tree species of Central Maine

Height-diameter models were developed for nine tree species common to the northeastern United States: Abies balsamea, Acer rubrum, Betula papyrifera, B. populifolia, Picea rubens, P. mariana, Pinus strobus, Populus tremuloides, and Tsuga canadensis. Stem heights and diameters were collected from 6 146 trees (between 136 and 2615 trees per species) on 50 plots within 10 structurally diverse stands that are part of a long-term silvicultural experiment in central Maine. The models were developed using both generalized nonlinear least squares (GNLS) and multi-level, mixed-effects approaches. Mixed-effects approaches were superior to GNLS, with inclusion of site covariates (tree density and basal area) accounting for some of the variability explained by the random coefficients in the full mixed-effect models. Analysis of plot-level parameter estimates suggested that differences in stand structure (even-aged vs. uneven-aged silvicultural practices) had a significant influence on the height-diameter relationships.