Satellite-observed mountain greening predicts genomic erosion in a grassland medicinal herb over half a century

Mountains are biodiversity hotspots contributing essential benefits to human societies, but global environmental change is rapidly altering their habitats. During the past five decades, increasing temperatures and land-use change in montane and subalpine elevations facilitated the productivity and expansion of competitive vegetation, termed as "greening," with adverse effects on open grassland habitats. Although vegetation greening is well-documented through satellite observations, its impact on the populations and genomic integrity of affected species remains underexplored. Here, we address this challenge by integrating 40 years of remote sensing data with museum genomics and fieldwork to assess the impact of mountain greening on the genomic diversity of grassland plants in the southern Balkan peninsula. We sequenced the genomes of historical and modern populations of Ironwort, a plant of significant medicinal value, and demonstrate widespread genomic erosion across its populations. Our results show that, on average, 6% (0%-20%) of Ironwort's genome is affected by inbreeding accumulation over the past half century, indicating various degrees of population declines. Importantly, we show that genomic erosion is highly predictable by the normalized difference vegetation index (NDVI) rates of change. Our models suggest that faster increases in vegetation density are associated with higher population declines in this grassland species, revealing the negative impacts of increasing productivity in mountain ecosystems. By linking two independent and disparate monitoring indicators, we demonstrate the ability of remote sensing to predict the consequences of environmental change on temporal genomic change in mountain grassland species, with far-reaching implications for protecting natural resources in these fragile ecosystems.