Assessing sources of variation on leaves reflectance spectra in coastal saltmarshes and seagrasses

There is an urgent need for effective large-scale biodiversity monitoring across ecosystems, given the recent tendency toward global biodiversity loss. The assessment of plant spectral diversity offers a promising approach as it is intrinsically linked to phylogenetic and functional diversity. This study investigates the relationship between taxonomic, functional, and spectral diversity in temperate saltmarsh and seagrass ecosystems in the Gulf of Biscay. Using hyperspectral leaf reflectance data and functional traits from 19 plant species across four estuaries, these three dimensions of biodiversity were compared. The predictive power of spectral data was assessed for estimating biochemical and anatomical traits and species identification. Results reveal significant correlations between functional and spectral diversity, with species sharing similar functional traits exhibiting similar spectral signatures. Spectral diversity is significantly influenced by taxonomic classification, with higher taxonomic levels (e.g., order, class) explaining substantial part of the spectral variation. Spectral regions of 720–770 nm and 1330–1380 nm were important for species discrimination, achieving 98% accuracy. Partial least squares regression models successfully estimated functional traits (e.g., water content, carbon, phosphorus) with high precision in these environments. These findings demonstrate that spectral data can effectively capture taxonomic and functional diversity, offering an effective tool for large-scale biodiversity monitoring in estuarine ecosystems. This study underscores the potential of remote sensing to track biodiversity and ecosystem health, providing a foundation for future applications in conservation and management.