Measuring the fractal dimensions of a temporal forest landscape

This study examines ambiguities in fractal analysis, a popular technique in landscape ecology used to quantify spatial patterns by computing the average shape complexity of patches. These ambiguities may arise depending on whether analysis is based on individual patches or on the entire landscape, and whether topographic influences are considered. Using data from the Olympic National Forest, Washington, annual fractal dimensions are examined for old-growth forest from 1900 to 1990. The conventional fractal measure of landscape structure is compared with results from both patch- and landscape-based analyses. In addition, topographic effects on fractal assessments are evaluated. Results show that the approach analyzing a landscape, in its entirety, may be a promising indicator for the structural complexity of landscape fragmentation, whereas topography appears insignificant to landscape fractal analysis. © 1998 John Wiley & Sons, Inc.