Modeling increased height, diameter, and specific gravity effects on yield estimates in planted loblolly pine stands

Increased growth of Pinus taeda L. plantations is expected from genetic improvement and intensified cultural practice. It is necessary for forest managers to have a means for estimating the yield of forest stands after increased growth occurs. Several models were tested to determine which could best predict dry weight yield after certain constant growth increases were assumed. The data used consisted of 189, one-tenth acre sample plots taken in loblolly pine plantations located in piedmont and coastal plain Virginia, and coastal plain Delaware, Maryland, and North Carolina. Constant growth increases in diameter, height, and wood density were considered. The models considered were multiple regression and the diameter distribution approach using the beta distribution. Under the diameter distribution approach alternatives involving a piecewise integration over diameter class limits and a "unified" approach were considered. The models were compared to each other and to expected yield increases when determining their consistency, accuracy, flexibility, and simplicity. Recommendations concerning the different models were made.