Predicting the Cover-up of Dead Branches Using a Simple Single Regressor Equation

Information on the effects of branch diameter on branch occlusion is necessary for building models capable of forecasting the effect of management decisions on tree or log grade. We investigated the relationship between branch size and subsequent branch occlusion through diameter growth with special attention toward the development of a simple single regressor equation for use in future hardwood stem quality models. Data were obtained from 21 boards representing 3 logs of the first 21 feet of one cherrybark oak originating from a planted stand north of Vicksburg, MS. Double cross-validation methods were used to evaluate fitted models. non-linear model form (Y = aBK max , where Y = overwood, BK max = maximum branch-knot diameter and a and b are parameters) provided the best fit. The model explained approximately 50 percent of the variation in overwood. INTRODUCTION Silviculturists have long realized the importance of tree or log grade. However, the implications of silvicultural activities on stem structure have been largely overlooked. This is particularly the case of recent large-scale replanting efforts in the Lower Mississippi lluvial Valley (King and Keeland 1999, Twedt and Wilson 2002), where many monospecific plantations lacking natural analogs are being created. Unlike some softwood products, grade production in hardwood trees is a more important factor in valuation than volume because of the great differential between the highest and lowest grades of lumber or veneer products produced. For example, the price differential between red oak F S and 1F alone and F S and 2 alone was 219 percent and 264 percent in March of 2005, respectively (Hardwood Market Report, 3/05/05). Therefore, understanding the impacts of silvicultural activities on the production of hardwood tree grade is critical. Experimental methods of acquiring causal information regarding the impacts of management activities on tree structure are needed, and in some cases are underway (Clatterbuck and others 1987, Oliver and others 1990). Complementary techniques that can expedite acquisition of needed information are necessary. Stem analysis techniques combined with modeling methods can improve our understanding in the interim, and help guide current and future land management decisions. s gross crown dimensions are proportional to and determinants of tree growth ( ssman 1970, Rennolls 1994), the number and size of branches within the crown are major determinants of stem structure and, therefore, wood quality. Wood quality is heavily affected by the development of first-order branches within the crown, particularly the self-pruning and subsequent occlusion of branches as crown recession occurs (Makinen 1998, Makinen and Colin 1998, Makinen and Makela 2003). Thus, a logical first step is to evaluate the effects of variable branch sizes on the stem diameter needed for branch occlusion. Information on the effects of branch diameter on branch occlusion is necessary for building models capable of forecasting the effect of management decisions on tree or log grade. However, little is known regarding the relationship between branch size and the occlusion of that branch through diameter growth following crown recession. The knowledge gap is particularly large for hardwoods, including highly valuable species such as cherrybark oak (Quercus pagoda Raf.). Models combining growth and development of stem structure, including internal characteristics, are in development (Maguire and others 1 Christopher M. Oswalt, Graduate Research ssistant, The University of Tennessee, Department of Forestry Wildlife and Fisheries, Knoxville, TN 37996–4563; Wayne K. Clatterbuck, ssociate Professor, The University of Tennessee, Department of Forestry Wildlife and Fisheries, Knoxville, TN 37996–4563; and E.C. urkhardt, urkhardt/Hardwood ssociates, Vicksburg, MS 39180.