Log-grade volume distribution prediction models for tree species in red oak-sweetgum stands on US mid-south minor stream bottoms
Red oak (Quercus section Labatae)-sweetgum (Liquidambar styraciflua L.) stands growing on mid-south bottomland sites in the United States are well known for producing high-quality grade hardwood logs, but models for estimating the quantity and quality of standing grade wood in these stands have been unavailable. Prediction models to estimate total merchantable sawlog volume and volume by grade category within a standing tree are presented and discussed. Nonlinear regression, binary logistic regression, and probability distribution function techniques were used in the development of individual tree models for species groups based on 2,149 professionally graded trees. Species, dbh, total height, and merchantable height were input variables used to determine projected volumes for each of five volume units and five species groups: cherrybark oak (Quercus pagoda Raf.), other red oak, all red oak, sweetgum, and other commercial species. Total merchantable sawtimber volume models by species group provided logical and consistent predictions with nonlinear R 2 values ranging from 0.70 to 1.00 depending on species group and volume unit. The estimated volume by grade category varies by tree but will closely estimate average grade volume when applied across numerous stands. The resulting equations were incorporated into a growth and yield simulator (software available from the Forest and Wildlife Research Center, Mississippi State University website). The ability to predict the quantity and quality of merchantable volume for these bottomland tree species should greatly advance the valuation of bottomland hardwoods and encourage the establishment and improved management of these stands.