Utilising Remote Sensing Technologies to Upscale Microbial Environmental DNA Point Data for Large Scale Biodiversity Assessment

Microbial communities are central to forest ecosystem functioning, acting as key drivers of ecosystem health, structure, and sustainability. Once largely overlooked, these communities are now increasingly incorporated into biodiversity assessments, driven by advances in environmental DNA (eDNA) techniques that enable monitoring of this previously invisible world. However, large-scale assessments of these communities remain limited as they require extensive sampling efforts to generate data with sufficient spatial coverage and resolution. To overcome this limitation, rapid advances in remote sensing technologies, combined with in-situ data and machine learning, are transforming biodiversity science by enabling consistent monitoring of ecosystem features from local to global scales. Here, we present a novel approach to modelling microbial diversity by coupling image spectroscopy data with eDNA profiling using machine learning models such as Gaussian Process Regression and Partial Least Squares Regression. By integrating point-based in-situ eDNA metabarcoding data (including open-access datasets) with Earth Observation data, we demonstrate that microbiome biodiversity can be reliably modelled, mapped, and evaluated across diverse ecosystems and spatial scales in European temperate forests. Our results show that hyperspectral remote sensing effectively captures habitat features that shape microbial diversity, including belowground communities. Additionally, we revealed that microbiological community structure signals underlying ecological relationships in temperate forests in Europe. Our methods enable scalable biodiversity and ecosystem monitoring and supporting research across broader taxonomic and spatial scales. By integrating the strengths of these state-of-the-art research fields, we can gain deeper insights into the spatial dynamics of microbial communities and their drivers of change. Due to the key role of microbial communities in ecosystem functioning, this interdisciplinary approach also opens new opportunities for predictive ecosystem monitoring and early detection of environmental degradation. By providing a scalable framework that enhances the integration of local observations within large-area assessments, this study supports more informed conservation and land management strategies to achieve regional and global biodiversity conservation targets.