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

Plants persist in variable environments as hosts to diverse bacterial, archaeal, fungal and protistan communities whose composition and activity can influence nutrition, development, immunity and stress tolerance. The holobiont concept captures this ecological interdependence, but its evolutionary interpretation remains contested because community membership, transmission fidelity and fitness alignment differ markedly among plant–microbe associations. This critical narrative review evaluates how coevolutionary processes, microbiome assembly and microbe-mediated adaptive plasticity intersect under contemporary environmental change. Literature published principally from 2010 to 15 July 2026 was selected from open scholarly databases and citation networks, with earlier foundational evidence retained where necessary. The synthesis indicates that plant microbiomes are assembled hierarchically from environmental and inherited source pools through compartment-specific filtering, host genotype and development, immune and nutritional signalling, exudate chemistry, dispersal, priority effects and intermicrobial interactions. Host genetic effects are reproducible but often modest relative to soil, site, age and weather, which constrains simple interpretations of microbiome heritability. Drought provides the strongest current evidence that environmental stress can restructure root microbiomes in recurrent directions and, in some systems, recruit or enrich microorganisms that improve subsequent plant performance. Yet stress-associated community change alone does not establish adaptive plasticity: convincing inference requires demonstrated host fitness benefits, environmental contingency, causal microbial manipulation and preferably reciprocal or multigenerational tests. Evidence for whole-holobiont coevolution is likewise weaker than evidence for coevolution within particular, persistent partnerships. Synthetic communities, culture collections, host genetics and multi-omics have strengthened causal analysis, although laboratory-to-field transfer remains limited by ecological context dependence. The review therefore supports a pluralistic plant-holobiont framework in which ecological integration is common, evolutionary integration is conditional, and adaptive microbiome functions emerge from testable interactions rather than from an assumed unitary hologenome.