Hyperspectral Airborne Remote Sensing Estimates Functional Diversity Rather Than Species Diversity at a Local Scale for Temperate Forests

Biodiversity is typically defined as the variety among and between species. For forest ecosystems, this definition has been expanded to include other aspects of diversity such as function, phylogenetics, and physical structure. These metrics of diversity are typically measured using field data, but advancements of remote sensing technology have increased the ability to estimate these metrics remotely. In this study, we tested the ability to estimate four metrics of biodiversity (species, functional, phylogenetic, and structural diversity) for a northeastern United States temperate forest, at a local scale, using airborne hyperspectral remote sensing. The four metrics of biodiversity were quantified using inventory data. Light Detecting and Ranging (LiDAR) data was used for structural measurements not included in the inventory assessment, such as height. Measures of vegetation indices, spectral variability, and mean reflectance were calculated from the hyperspectral imagery and compared with the four biodiversity metrics. The ability of canopy reflectance data to estimate diversity metrics was tested using simple linear regressions and Partial Least Squares Regression (PLSR) models. Vegetation indices and spectral variation were found to be weakly correlated to all biodiversity metrics at this local scale, but mean reflectance was strongly correlated to functional diversity. PLSR models were able to explain 54% of the variation of functional diversity. Leaf based traits, such as foliar nitrogen, leaf type, and shade tolerance, contributed the most to this relationship and were estimated most strongly by wavelengths in the near-infrared (NIR) and short-wave infrared (SWIR) regions. Therefore, at a local scale, functional diversity based on leaf traits can be estimated remotely using hyperspectral imagery and that the variation in functional traits are driving reflectance patterns more than the variation of the species themselves.