Methylated memories mechanism, evidence, and skepticism in plant transgenerational epigenetic inheritance

Plants are repeatedly offered as the strongest case that organisms can inherit acquired experience without a change in DNA sequence - a claim that brushes against more than a century of Weismannian orthodoxy. DNA methylation, small-RNA pathways, and chromatin marks can be reshaped by heat, drought, salinity, and pathogen attack, and in several plant systems these marks persist into unstressed offspring. Yet the field remains split. Well-controlled studies find epigenetic memory to be shallow, genotype-dependent, and easily confused with parental provisioning or incomplete resetting during gametogenesis, while a smaller set of striking systems - among them paramutation at the maize b1 locus and stress-induced methylation in clonal dandelions - show that heritable, sequence-independent memory is mechanistically real, at least in specific genomic contexts. This review focuses on DNA methylation, the best-characterized carrier of heritable epigenetic information in plants, and in particular on RNA-directed DNA methylation (RdDM), the de novo pathway most closely linked to stress, while placing it alongside the MET1 and chromomethylase pathways and chromatin-based memory. It uses maize paramutation and dandelion stress memory as two contrasting, illustrative case studies, situates them within the wider body of stress-methylation experiments in Arabidopsis, rice, and other species, lays out the skeptical counter-case, and considers why commercial interest in "climate-smart" epigenetically primed seed is forcing the field to resolve these questions faster than the data currently allow.