The basis for the selection of trees to be used in the production of dendrochronologies has long been an issue (Douglass 1946; Fritts 1976). In humid regions the common practice has been to use trees that appear to be in good health. As a part of a larger study involving the impact of ice storms on tree-ring increments (Travis 1989), we show that trees stressed as a result of ice damage produced a stronger climatic signal than nondamaged trees.

The effect of ice damage on loblolly pine (Pinus taeda) plantations in the Piedmont region of Virginia was examined to discover if management decisions can minimize net present value losses. A simulation approach was used for the analysis. Loblolly pine plantations were simulated using the growth and yield model, Trulob. Ice damage was factored into the model using prediction equations from a previous study. A decision tree framework was used to determine which management plans resulted in the highest net present value. The results show that ice storms can cause significant losses to the net present value of loblolly plantations. In most instances changing management plans could not minimize losses. In situations where altering management plans can result in higher net present values if ice occurs, landowners should also be aware of the suboptimal net present value they will be returned if these plans are followed and no ice storms occur. / Master of Science