State University of New York, College of Environmental Science and Forestry, Syracuse, NY, 12310, USA

Background and objectives
Emphasis on environmental awareness in recent decades has resulted in increased forest health monitoring activities throughout the world, but interpretation of forest health or lack of health involves much more than quantifying changes in crown condition, disease incidence, and mortality. Proper and sustained forest development needs "a healthy amount of disease" and therefore forest health assessment should be based on realistic expectations [1] derived from appropriate stand structure models such as the law of de Liocourt [2] expressed by ln N = aX + c, where N is the number of trees per unit area in each diameter class, X; the slope of the line, a, is the relative survival rate per diameter class, and c is the number of trees in the zero diameter class. Expected relative mortality, m, is calculated as m = 1 - eDa, where D is the width of the diameter class. Objectives are to evaluate this sustained life from death (phoenix helix) model based on a large forest system made up of a number of forest types.

Materials and methods
Data on 2650 trees from 201 variable radius plots at 67 random points in the Adirondack Park of New York State were collected in the summer of 1996. Mortality and health liabilities were summarized by diameter classes within forest types. Branch, foliage, stem, and root defects were given numerical weighted scores. Relative crown position, ratio of live crown to total height, and breakage were also given weighted scores. These two groups of weighted scores were summed to generate health liability ratings for individual trees, stands of trees, or species of trees.

Results and conclusions

Numbers of trees within diameter classes fit the natural log function, described above, for the total forest and for each of the ten forest types within. Expected relative mortality for trees to grow 2.5 cm for any diameter class of trees in the Adirondack Park was between 19 and 26% depending on the forest type. The actual relative mortality in the sample was generally below expected for trees less than 50 cm. Traditional interpretation of these data might suggest that the forest is unhealthy since there is currently 20% mortality in this forest. Simple interpretation of the data suggests that the forest is healthy because there is less than expected mortality. Analysis based on stand structure, following the law of de Liocourt, suggests that this healthy- looking forest is out of balance, and that some biological or abiotic disturbance will be needed to return the system to equilibrium. Some excess mortality, greater than expected, is readily explained, e.g. excess mortality in large beech is from beech bark disease and excess mortality in mid-size red spruce and balsam fir is from a recent blow down. However, excess mortality in large yellow birch is unexplained. Interpretation of liability factors provides some indication of the nature and abundance of future mortality balancing factors. These observations suggest that assessment of forest health based on the phoenix helix interpretation of forest stand structure provides a realistic foundation for addressing forest health issues.

1. Manion PD, Griffin DH, 1997. Phytopathology 87 (supplement), S62.
2. Meyer HA, 1952. Journal of Forestry 50, 85-92.