Large landscape scale analysis of tree death in the Adirondack park, New York

We propose that healthy forests depend on quantitatively predictable tree death as a continuous process linked to forest structure and growth. Quantifying a baseline mortality value by forest structure analysis using the Law of de Liocourt allows estimation of forest health by comparing the observed mortality to a baseline value. This method is applicable to large landscape samples, but not generally to individual forest stands. Observed relative mortality per dbh class from a random sample of the Adirondack Park (New York, USA) forest was slightly, but significantly, less than the baseline relative mortality per dbh class required for maintaining size-distribution stability of the forest. This result suggests a changing structure involving increased forest density, or future mortality increase to maintain the current structure. Differences in structure-mortality relationships among the more abundant species indicate changing composition of the forest. This reflects unhealthy conditions in some species (American beech, yellow birch, balsam fir, and red spruce) that are compensated by enhanced development of others (red maple, sugar maple, and eastern hemlock). For. Sci. 47(4):542–549.