Spaceborne remote sensing effectively maps species richness across taxonomic groups in a mountain landscape

Biodiversity decline due to global change poses a pressing challenge for conservation efforts worldwide. To improve the efficiency of conservation projects, spatially explicit information on species richness is needed, yet this information is challenging to generate from traditional biodiversity assessments. To fill this gap, we here explored the potential of spaceborne remote sensing techniques, including Sentinel-1, Sentinel-2 and EnMAP, for mapping species richness across four distinct taxonomic groups (fungi, plants, insects and birds) in a complex mountain landscape in the German Alps. We used all sensors individually, as well as different combinations of data (Sentinel-1/2, EnMAP/Sentinel-1/2), and compared predictions to predictions based on LiDAR data – a well-proven standard in mapping species richness. Our results showed that a combination of EnMAP/Sentinel-1/2 performed as well or even better than airborne LiDAR data for predicting species richness, but predictive accuracies of individual spaceborne models were substantially lower. This suggests that optical, radar and hyperspectral data carry complementary information and combining this information unleashes the full potential of spaceborne data for species richness mapping. However, validating models by habitat type revealed higher errors within habitat types (i.e., forest or open habitat), especially for immobile species (fungi and plants) that likely vary at smaller spatial scales than the resolution of the spaceborne systems used in this study. Overall, our findings highlight the potential of spaceborne remote sensing for large-scale biodiversity assessments, offering valuable insights into spatial biodiversity patterns and their changes over time.