Can plants pass stress memories to their offspring? The evidence is mixed

Where stress marks persist

The review concludes that plant epigenetic inheritance is neither an established general rule nor an illusion. In specific cases, molecular marks can persist through cell divisions and, at some loci such as maize b1, through meiosis. The review identifies RNA-directed DNA methylation as a plausible route by which stress-linked small RNAs can install new DNA methylation, while maintenance pathways involving MET1 and chromomethylases can help preserve methylation.

The dandelion examples show that ordinary environmental stresses can alter methylation patterns that partly persist in unstressed offspring and are accompanied by measurable phenotypic effects. Across the broader Arabidopsis and rice literature discussed in the review, however, most induced changes are described as short-lived and reset within one or two generations. The authors distinguish three questions: whether a mark is inherited, whether it benefits the offspring, and whether the pattern is widespread. They find strong evidence for inheritance in some settings, mixed evidence for benefit, and weak evidence for generality.

Why the limits matter

The findings place limits on the idea that plants broadly pass on acquired stress experience. They also matter for interest in “climate-smart” or epigenetically primed seeds: a methylation change that persists in one plant system may not behave the same way in a major crop or in plants that reproduce sexually. The review argues that future work and commercial claims need to test inheritance, benefit and generality separately rather than treating epigenetic memory as a yes-or-no trait.

Evidence and caveats

This is a review of published research, not a new experiment. Its strongest examples come from particular systems: maize paramutation and clonal dandelions. The dandelion system reproduces without meiosis or fertilization, bypassing much of the resetting that occurs in sexually reproducing plants; the review therefore treats its transmission rates as an upper bound rather than a typical value.

The authors also note that stress-related methylation can be confused with parental provisioning or incomplete resetting during reproduction. Results vary by species, genetic background, genomic location and environment. Because most induced changes in the broader Arabidopsis and rice literature are short-lived, findings from clonal dandelions cannot simply be extended to outcrossing or self-fertilizing plants, including most major seed crops.