We now know that heritable phenotypic variation within species can not only be caused by underlying genetic variation, but also by heritable variation in epigenetic modifications of the genome, such as DNA methylation. However, in most systems it is impossible to disentangle genetic and epigenetic effects, and to unambiguously demonstrate the phenotypic consequences of epigenetic variation. A rare exception are epigenetic recombinant inbred lines (epiRILs) of Arabidopsis thaliana, which are genetically identical but significantly different at the level of DNA methylation. We used these epiRILs to thoroughly test for the effects of epigenetic variation on complex plant traits and the plasticity of plants to different environmental changes. In a common garden, we grew 477 epiRILs with replicates in with and without intra-specific competition treatments. In another greenhouse experiment, we planted replicates of 134 epiRILs as well as several control lines into three different environments (control, drought stress and nutrient addition). We examined heritabilities on means and plasticities of ecological important traits (flowering time, root: shoot ratio, biomass, reproductive allocation, fitness, etc.)
We found significant heritability in all of these traits, and we also found significant variation in plasticity among the epiRILs in response to intra-specific competition, drought stress and nutrient addition. Heritabilities of trait means (H2) ranged from 0.047 to 0.357 were generally larger than heritabilities of plasticities (HI2) expect for root: shoot ratio, HI2 of which in response to drought was much higher than H2. Flowering time seemed would undergo balancing selection. Moreover, we found plasticities of decreasing plant height and increasing root: shoot ratio in response to drought was adaptive. As there was also significant heritability of traits and plasticities, the epiRIL population should be able to evolve rapidly only on the basis of epigenetic variation.