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

Plant-associated microbiomes can alter nutrient acquisition, defence, development and tolerance to abiotic stress, yet their contribution to plant resilience under global change is frequently inferred from taxonomic shifts rather than demonstrated mechanistically. This critical narrative review integrates evidence on how drought, warming, elevated atmospheric carbon dioxide, salinity and interacting climate drivers restructure plant microbiomes, and asks when those changes constitute ecological resilience, eco-evolutionary feedback, stress memory or heritable microbial contribution across generations. Literature published from 2010 to 15 July 2026 was considered, with earlier conceptual evidence retained only when necessary. Evidence was appraised for ecological realism, temporal resolution, causal identification, host and soil context, functional validation and capacity to distinguish plant, microbial and environmental legacies. The strongest cross-system signal concerns drought-associated bacterial reassembly, including recurrent enrichment of monoderm lineages, but conserved taxonomic responses do not consistently predict plant-beneficial function. Long-term and repeated-stress experiments show that microbial communities can retain drought legacies that alter subsequent ecosystem or plant responses, although such effects are contingent on host identity, soil history and experimental scale. Experimental-evolution studies establish that microbes can evolve rapidly enough to modify host fitness, but direct demonstrations of reciprocal eco-evolutionary feedback in complex field microbiomes remain scarce. Seed and endophyte studies provide convincing evidence for transmission of selected microbial members across plant generations; evidence that stress specifically reprogrammes those transmitted communities to increase descendant resilience is much weaker. A central synthesis is therefore that microbiome-mediated resilience should be evaluated as a process linking disturbance, community or trait change, persistence or recovery, and measurable host performance, rather than as compositional stability alone. Progress will require factorial multigenerational experiments, strain-resolved tracking, multi-kingdom and functional measurements, reciprocal microbiome transfers, and field tests that separate host genetic or epigenetic inheritance from microbial transmission. Plant microbiomes are plausible contributors to adaptation under global change, but their predictive and heritable roles remain conditional rather than universal.