Details

Autor(en) / Beteiligte

• Brown, J.P
• Sendak, P.E

Titel

Association of ring shake in eastern hemlock with tree attributes

Ist Teil von

• Forest products journal, 2006-10, Vol.56 (10), p.31-36

Ort / Verlag

Madison, WI: Forest Products Society

Erscheinungsjahr

2006

Quelle

Access via Business Source (EBSCOhost)

Beschreibungen/Notizen

• Eastern hemlock (Tsuga canadensis (L.) Carr.) is a major or minor associate in many forest types in northeastern North America. There has never been the high level of demand for eastern hemlock lumber that characterizes competing softwoods like white pine and spruce. Nevertheless, periodically interest in greater utilization of eastern hemlock for lumber has encouraged attempts to solve some of the wood utilization problems associated with the species. Ring shake, the longitudinal separation of wood parallel to the growth rings, is one of the more serious problems affecting utilization of hemlock lumber. Ring shake is not allowed in the better and more valuable grades of hemlock boards and it also reduces the grades of dimension lumber. Based on a large sample of trees widely distributed over hemlock's range in the Northeast, we determined whether any tree characteristics are prominent indicators of ring shake in hemlock lumber and estimated shake volume in lumber from trees that were identified as likely to contain shake. A logistic regression (LR) model was used to determine significant factors affecting shake presence (shake percent > 0). For trees recorded as having shake, an ordinary regression model that estimated shake volume was developed. The best LR model to determine presence of shake included the significant predictors bird peck, age, and DBH (a = 0.05). This model correctly predicted 77 percent of the validation cases. The best regression model to estimate shake volume in trees having shake included just one significant predictor: DBH. Our results indicate that both older and larger trees are more prone to shake, as well as trees exhibiting bird peck. The ordinary regression equation was only moderately successful at predicting the amount of shake volume.