Spatial Patterns of Genomic Variation and Genomic Offset in a Common Grassland Plant and Their Relation to Seed...

Restoring temperate grasslands often necessitates the introduction of large quantities of seeds, a process that is regulated by seed transfer zones in many countries. These zones are commonly delineated based on abiotic factors. Consequently, it remains uncertain to what extent existing seed zones represent and thereby protect or erode the spatial distribution of genetic variation. Empirical data on the spatial genetic structure of grassland species are therefore essential to address this knowledge gap. Moreover, as seed zones are increasingly expected to provide genotypes pre-adapted to climate change, such data can also inform predictions of maladaptation and support the identification of suitable donor populations. Here, we focus on Galium album, a widespread perennial grassland species, which we sampled systematically across Germany, with an average of one population per 25 × 25 km area. Based on 8348 SNP loci, we analyzed the population genetic structure using Bayesian clustering. We identified four spatially coherent genetic clusters, which explained 2.43% of genomic variation but showed little congruence with current seed zones. Yet, seed zones still capture a significant component of spatial genetic structure (1.92%), which is also reflected in a significant isolation by distance among zones. Seed transfer practices are increasingly challenged by climate change, shifting the adaptive requirements for populations. We performed a genotype-environment association analysis using redundancy analysis, and estimated the genomic offset, that is, the genomic change required to maintain the current genotype-environment relationship under climate change. The genomic offset was generally moderate across Germany, even under a pessimistic climate scenario projected into the more distant future (SSP5-8.5, 2081-2100). For one of the few locations where the temporal genomic offset slightly exceeded a previously proposed threshold, we identified suitable donor regions harbouring potentially pre-adapted genotypes for targeted assisted migration, both within the same and in adjacent zones.