Numerous studies have shown the capacity of imaging spectroscopy data to map a range of foliar traits that describe the function, diversity and phylogeny of vegetation. This “spectranomics” approach (Asner & Martin 2016) offers the potential to rethink how we quantify ecosystems and gap-fill our knowledge of plant function and biodiversity, especially in locations where field measurements are sparse. In advance of SBG and other global-scope hyperspectral satellites this decade, there is now interest in applying the trait-mapping across ecosystems globally. Yet, challenges remain as we test our capacities using airborne imagery. First, the methods to develop these products and generality of those methods have not been established. As well, we must identify how the results of imaging spectroscopy will get translated into information both to supplant traditional “plant functional type” (PFT) approaches used by most earth system models (ESMs) and to explicitly represent functional diversity within those models. Finally, the vast data on functional traits we are now generating from hyperspectral imagery is actually transforming our understanding relationships among traits (and their variation with environment), since we can now concurrently measure many more traits over much broader areas than using traditional field sampling and lab assays. Here we demonstrate the new insights being provided by trait mapping from multiple imaging spectroscopy studies.
Results/Conclusions
Functional diversity in ecosystems sometimes shows up where you least expect it. First, the trait-vs-trait relationships that are typically used to characterize ecosystem variability (e.g., the leaf economics spectrum) are not always the most interesting discriminators of spatial variation in functional traits. Second, results from analyses of imagery show that some ecosystems that are low in species diversity (e.g. the Arctic) may actually be highly functionally diverse. Conversely, comparatively higher species diversity systems may not always show up as such in interpretations drawn from imagery, e.g. in forests where only crowns are measured by imagery. As well, the timing of imagery and phenological variation in traits can strongly influence the resulting interpretations of function and diversity. Finally, we find that there can be high levels of functional diversity in semi-natural (or managed) ecosystems and in agriculture. This will dramatically affect how we include such ecosystems in the next generation of ESMs. Results from a range of studies including NEON, ABoVE and other NASA studies show that generalization is possible but the interpretations will vary across environmental and geographic space.