Effects of thinning an ring specific gravity in loblolly pine trees were examined in a regional trial in the southeastern United States. The thinning study consisted of two treatment levels (30% and 50% of the basal area removed) and a control. Thinning did not alter the proportion of the annual ring in earlywood or latewood and thus did not affect ring specific gravity. There was regional variation in average ring specific gravity, with trees from the Coastal Plain having higher ring specific gravity than trees from the Piedmont or the Highlands. Within the Coastal Plain, trees in the Gulf Coastal Plain tended to have higher average ring specific gravity than trees in the Atlantic Coastal Plain. A ring specific gravity prediction model was developed based on ring position, tree, stand, and site attributes. Much of the variation in ring specific gravity was due to within-tree variation; consequently, stand and site factors accounted for a limited proportion of variation in ring specific gravity, As the within-tree observations used in the study tended to be correlated, a direct covariance modeling approach was used to account for the correlation among observations.
