Thu, Aug 18, 2022: 5:00 PM-6:30 PM
ESA Exhibit Hall
Background/Question/Methods: Ecological diversity has been shown to strongly influence ecosystem functions, processes, and provision of services to humans. The estimation of biodiversity using remote sensing technology is increasingly important due to the unpredictable effects of climate change. Measuring diversity over large spatial and temporal scales necessary to monitor for the effects of climate change is challenging and resource-intensive. The spectral variation hypothesis postulates that spectral diversity of vegetation is a direct link to species richness and other forms of spatial diversity (Palmer, 2002). We developed methods to estimate diversity and composition of natural forest canopies from airborne imaging spectroscopy. This research uses hyperspectral data from one image acquired in July, 2019 (400-1000 nm, 178 bands, 1 m GSD) to calculate spectral diversity. In this study, we analyze variation in spectral composition and link this to ground-based estimates of species richness, evenness, and beta diversity in a deciduous, broad-leaved forest in Nebraska. We used detailed tree inventory data from the Indian Cave forest plot that was collected contemporaneously with the image, and apply spectral diversity and composition analyses across management units that have received different histories of prescribed burning and thinning.
Results/Conclusions: We have integrated established techniques to calculate spectral diversity over forest canopies by means of hyperspectral data. We used a multiple linear regression analysis over spectral diversity and previously mentioned variables of ground-based measurements. We show that spectral diversity can follow a well-defined trend to ground-based richness measurements where higher spectral variation increases as a combination of species richness and evenness does as well. We observe that spectral diversity among communities shows some positive trend with beta diversity, however these measurements seem to depend on defined community extents. For management units, our results show some possible correlation where higher spectral diversity is associated with lower intensity treatments or areas that have experienced a longer time lapse since treatment, or a combination of both. Management unit results vary more than those of spectral diversity and ground-based measurements. This study demonstrates the potential of remote sensing in monitoring biodiversity over large scales, supplementing traditional data collection and enabling targeted management and conservation efforts.
Results/Conclusions: We have integrated established techniques to calculate spectral diversity over forest canopies by means of hyperspectral data. We used a multiple linear regression analysis over spectral diversity and previously mentioned variables of ground-based measurements. We show that spectral diversity can follow a well-defined trend to ground-based richness measurements where higher spectral variation increases as a combination of species richness and evenness does as well. We observe that spectral diversity among communities shows some positive trend with beta diversity, however these measurements seem to depend on defined community extents. For management units, our results show some possible correlation where higher spectral diversity is associated with lower intensity treatments or areas that have experienced a longer time lapse since treatment, or a combination of both. Management unit results vary more than those of spectral diversity and ground-based measurements. This study demonstrates the potential of remote sensing in monitoring biodiversity over large scales, supplementing traditional data collection and enabling targeted management and conservation efforts.