THE SPATIAL DISTRIBUTION OF FOREST COMMUNITY ABUNDANCE DISTRIBUTIONS IN THE EASTERN UNITED STATES
Species rank abundance distributions (RADs) demonstrate a near universal pattern; there are very few abundant species while there are many species that are extremely rare. By matching these distributions to known biological distributions, researchers may be able to discern how communities partition resources and niche space. Whether or not these models have a broad spatial pattern is something that has yet to be determined. Using the Vegan package found in R Statistical Software, twenty-three forest inventories found across the eastern United States were analyzed to determine the best fit RAD for each inventory. A bootstrapping technique was used to determine whether the best fit model was consistent over various sample sizes for each inventory. Calculated distribution parameters for the best fit RAD were regressed against a spatial surrogate to determine the presences of a spatial pattern. Large scale RADs were calculated for the eastern U.S. using the U.S. Forest Inventory and Analysis plots. Both model type (qualitative) and model distribution parameter (quantitative) were mapped using a GIS.
Andrew Trgovac - University at Buffalo - SUNY
Dr. Chris Larsen - University at Buffalo - SUNY