Determinants of tree quality and lumber value in natural uneven-aged southern pine stands

An ordered-probit model was developed to predict tree grade from tree- and stand-level variables, some of which could be changed by management. Applied to uneven-aged mixed loblolly (Pinus taeda L.) - shortleaf pine (Pinus echinata Mill.) stands, the model showed that the grade of pine trees was highly correlated with tree diameter, tree height, and stand basal area, in non-linear fashion. In addition, a tree was more likely to be of high quality if it grew on industry or government forestland, on poorer sites, and in stands that had been partially cut in the past. However, the effects of changes in these variables on the unit value of recovered lumber were small. The exceptions were tree diameter and height, which were the most important indicators of lumber value.