Unraveling the seasonality of functional diversity through remote sensing

Abstract Plant functional diversity is a key biodiversity indicator of ecosystem dynamics. Several studies have shown that spectral data from Earth observation satellites will enable us to map functional diversity over large areas. However, most studies only consider snapshots of such data in time, and our knowledge of the temporal variation of functional diversity across the World’s biomes remains sparse. Here, we use hyperspectral remote sensing and deep learning to explore multi-seasonal functional diversity patterns on a global scale. We show that functional diversity can be highly dynamic over time. These dynamics are biome-specific, driven by seasonal cycles and wet-dry periods. Our findings highlight that the role of functional diversity in shaping ecosystem dynamics can be unraveled only by incorporating seasonality into functional diversity mapping. Such a multi-temporal approach will enhance the robustness of functional diversity assessments and deepen our understanding of ecosystem responses to environmental change.