New York State Forest Health Monitoring Project

Naja Kraus  and  Ben Rubin

Western New York -- 2000

BACKGROUND
The USDA Forest Service National Forest Health Monitoring (FHM) Program is currently in process of being implemented in New York State. In addition to forest health monitoring activities being carried out at selected, pre-established Forest Service Forest Inventory Analysis (FIA) plots, the program provides for an independent network of plots to more completely cover the land area of the State. This was an excellent opportunity for a group of professors and graduate students in the EFB Faculty who are already interested in forest health monitoring. We have developed a study, built on some procedures and hypotheses that grew out of earlier forest monitoring projects of the Adirondacks and northern New York, that is designed to meet the needs of the National FHM Program in New York. During the first of three years, we sampled the western third of New York State (NYS-DEC regions 7, 8 and 9). We plan to sample southeastern and northeastern New York over the next 2 summers.

DEFINITION
The health of a forest is its ability to sustain itself if left undisturbed, or to recover rapidly if it is disturbed. Forest health is measured on a relatively large spatial scale. Although individual stands within an area may be undergoing change or disturbance, they do not significantly impact forest health if they are compensated for by complementary changes in other stands.

OBJECTIVES
Our objectives are to complete a sampling of the health or sustainability of forested, state owned-land in New York over the course of three years. Specifically, we will:

1) look for potential threats to forest health from:
· unbalanced distributions of large and small trees, such that regenerating trees are not    numerous enough to eventually replace mature trees at the current level.
· mortality and / or cutting rates that are too high or too low to allow mature trees to be    replaced at the current level.
· forest disturbances due to severe weather conditions or forest insect or disease outbreaks.
· invasive species that may alter the habitat for forest plants and animals in the future.

2) examine the relationship between biodiversity and forest health
.
3) classify our plots by forest community type, so that we can analyze the health of individual forest types found within New York.

Furthermore, we hope that our dataset will also be useful in the future as a baseline against which future changes in the forest health and structure can be measured.

METHODS
Random points were located throughout the State Forest lands of New York. From May to August, 2000, crews of two or three monitors surveyed 163 locations. At each location we sampled trees equal or greater than 9 cm diameter at breast height (dbh) on three 10 BAF (basal area factor) prism plots and saplings (1.5 cm to 9 cm dbh), herbaceous and shrubby vegetation on nine 10 m^2 subplots. In addition, we collected data on the environmental conditions at the site including topographic, drainage, soil and land use history information, as well as digital photographs of the canopy structure. Each plot was marked so that it can be resampled in the future.
      For each tree sampled we recorded species, dbh, and information on whether it was alive or dead, on its physical structure, on its biotic or abiotic disease problems (if any), and on its crown characteristics.
     For understory vegetation we measured percent cover by species.

ANALYSIS
Obviously, the scope of our objectives and dataset require many different types of analysis that are currently ongoing. One simple example is a description of the species composition of the forest in terms of importance value (IV). IV is an index, which synthesizes the basal area, density and frequency of occurrence of each species (see figure). A more complex analysis is based on a conceptual model we have developed called the Phoenix helix concept. The basic idea is that as populations of trees grow, their numbers must decrease. Like the legendary bird, the forest derives life from death. Unlike the one for one replacement of the Phoenix, the forest forms an upwardly spiraling population of increasingly larger but fewer trees. Therefore, there must be a level of mortality in the forest that is considered 'normal ' or 'baseline' mortality. To estimate this rate, we record trees by diameter classes and compare the number of trees per hectare in each diameter class.
     In the past we have found that if we plot the diameter distribution on a logarithmic scale, it forms a linear pattern. Based on a regression line fit to the diameter distribution, we can calculate the percent decrease from one diameter to the next. We refer to it as the 'predicted mortality' and consider it an estimate of the baseline mortality predicted by the Phoenix helix concept. In many populations the percentage of trees in each diameter class that are dead is approximately equal to the constant predicted mortality. We interpret this as a balanced or healthy population structure. For 2.5 cm (1") diameter classes, the density of trees in any diameter class is approximately 20% to 25% less than that of the previous diameter classes. Furthermore, this percentage is constant throughout most diameter classes for almost all large populations, indicating that every time trees in the population grow by one diameter class a constant portion of them (between 20% and 25%) must die in order to sustain the population structure.
     Our current dataset offers some new possibilities. Although a regression line explains most of the variation in the diameter distribution (R^2 = 0.98), it appears that there is some systematic deviation from that line between diameters of 5 cm and 50 cm. It also appears that the observed mortality also varies over the same range. Our current questions are: Is the diameter distribution significantly different from its regression line? If so, why? Does the variation in observed mortality compensate for the difference between the predicted and observed density or, is the diameter distribution unsustainable indicating future changes in the forest's structure? Which forest community types are affected by the potential deviation from sustainability?


For more information contact:

Naja Kraus  nakrause@syr.edu      or     Ben Rubin   bdrubin@syr.edu