The Plant Microbiome in a Changing World: Ecology, Evolution, Stress Resilience and Sustainable Applications

Plants rarely experience environmental stress as isolated events. Drought, salinity, nutrient limitation and pathogen attack recur across seasons and generations, creating the possibility that previous exposure alters responses to later stress. Plant stress memory is usually discussed in terms of host signalling, transcriptional reprogramming and epigenetic states, yet plants also modify and inherit microbial communities whose responses can persist after the initiating stress has ended. This critical narrative review evaluates when such persistence can reasonably be described as microbiome-mediated stress memory, how microbial legacies arise, when they prime rather than merely accompany plant responses, and whether they contribute to resilience across plant generations. Literature was critically selected from major open scholarly databases and citation networks with emphasis on experiments that separated prior from current stress, manipulated microbiota or soil history, and measured plant functional outcomes. The strongest evidence comes from drought-conditioned soil and root microbiomes and from disease-suppressive plant-soil feedbacks. These systems show that stress can leave persistent microbial compositional, functional and chemical legacies that alter later plant performance. Nevertheless, beneficial outcomes are not universal: microbial legacies can be neutral or detrimental, depend on stress intensity, soil, genotype and host evolutionary history, and can be confounded by persistent abiotic soil changes. Mechanistic studies implicate host exudation, nutrient and immune signalling, microbial competition, osmotic and metabolic functions, and stress-conditioned community assembly, but recent work challenges simple adaptive “cry-for-help” interpretations. Microbial transmission through seeds and persistent soil legacies provide plausible routes of ecological inheritance, whereas evidence that stress-specific, adaptive microbiome information is reliably transmitted across multiple plant generations remains limited. Progress requires reciprocal transfer experiments, abiotic-legacy controls, strain-resolved and absolute-abundance measurements, multigenerational common-garden designs and field validation. Microbiome-mediated stress memory is therefore best treated as a conditional ecological process whose adaptive value must be demonstrated rather than assumed.