PARENT SESSION
Poster Session # 14: Forest Ecology.
Wednesday, August 6 Presentation from 5:00 PM to 6:30 PM. SITCC Exhibit Hall B.
Concisely characterizing forest dynamics with change vectors.
Fulton, Mark*,1, Harcombe, Paul2, 1 Bemidji State University, Bemidji, Minnesota, USA2 Rice University, Houston, Texas, USA
ABSTRACT- Changes in a forest stand or a set of monitoring plots can be envisioned as vectors in a multidimensional space defined by the abundance of tree species. These vectors can be used to concisely characterize aspects of forest chance that are more difficult to recognize in the trends of individual species. A vector of change is characterized by a direction and a magnitude. The direction of a change vector is determined by relative amount of change in each species; the magnitude is the total amount of change. One can compare directions of change among several monitoring plots by comparing angles between change vectors, either pairwise between individual plots or between individual plots and the overall trend. If one plot is monitored over an extended period of time, one can compare directions or magnitudes of change from one survey interval to the next to reveal the effects of changing conditions. Randomization tests can be used to detect consistent temporal trends in response to past disturbance or changing climate, thereby distinguishing trends from fluctuations of an essentially stable forest. Plotting changes in ordination diagrams has been used in the past to approach these kinds of questions, but comparing change vectors directly with each other avoids the distortions that are inevitable when a high-dimensional space is projected onto a low-dimensional space. We demonstrate this approach using long-term data from two stands in eastern Texas: a mesic forest disturbed by a moderate hurricane, and a bottomland hardwood forest with a variable flooding regime.
Key words: change vectors, stand dynamics, forest