Plant species diversity and composition of experimental grasslands affect genetic differentiation of Lolium perenne populations

Contrasting hypotheses exist about the relationship between plant species diversity and genetic diversity. However, experimental data of species diversity effects on genetic differentiation among populations are lacking. To address this, Lolium perenne was sown with an equal number of seeds in 78 experimental grasslands (Jena Experiment) varying in species richness (1, 2, 4, 8 to 16) and functional group richness and composition (1-4; grasses, legumes, small herbs, tall herbs). Population sizes were determined 4 years after sowing, and single-nucleotide polymorphism (SNP) DNA markers based on bulk samples of up to 100 individuals per population were applied. Genetic distances between the field populations and the initially sown seed population increased with sown species richness. The degree of genetic differentiation from the original seed population was largely explained by actual population sizes, which suggests that genetic drift was the main driver of differentiation. Weak relationships among relative allele frequencies and species diversity or actual population sizes, and a positive correlation between actual population sizes and expected heterozygosity also supported the role of genetic drift. Functional composition had additional effects on genetic differentiation of L. perenne populations, indicating a selection because of genotype-specific interactions with other species. Our study supports that genetic diversity is likely to be lower in plant communities with a higher number of interspecific competitors. Negative effects of species richness on population sizes may increase the probability of genetic drift, and selection because of genotype-specific interactions depending on species and genotypic community composition may modulate this relationship