PARENT SESSION
Oral Session #34: Forests: Models of function and structure. Presiding: M. Lechowicz.
Tuesday, August 7, 2001. 1:00 PM to 3:45 PM. Hall of Ideas F.

Using fractal geometry to estimate forest canopy structure.

Hartvigsen, Gregg¹, ¹

ABSTRACT- The three-dimensional structure of forest canopies is complex if details at the scale of leaves and branches are included. Structure at larger scales, however, can be easily estimated if we assume trees take on relatively simple, filled geometric shapes. I make the assumption that trees can be classified as ellipsoids, cones, or cylinders and investigate differences among two forests in upstate New York using fractal geometry. The box-count technique to determine fractal dimension allows quantitative comparisons of the shape of different forest canopies because the metric is scale-invariant, meaning that it is insensitive to differences such as the height of trees. I sampled two 20 x 20 meter plots in each of two forests of different developmental stages and found that the structure between these canopies differed significantly. These observed structural differences were robust, being unchanged by a randomization procedure that shuffled canopy shapes of individual trees. I suggest that this relatively simple method of assessing canopy structure is useful to investigate how plant canopies vary along environmental gradients, how forests of different developmental histories and stages differ structurally, and how these differences might influence other organisms such as bird communities.

KEY WORDS: fractals, forest, canopy