Abstract
BACKGROUND: Plants evolve as holobionts, ecological and evolutionary units made up of the host plant and its associated microbiota, which shape plant fitness and adaptive capacity. Isolated ecosystems with low biodiversity and plant cover, such as the fellfields of the remote sub-Antarctic Kerguelen Islands, represent ideal open-air laboratories to disentangle the drivers affecting plant-microbiome interactions. In such pristine environments, endemic plant species and their microbiota have coevolved in isolation possibly since the last ice age. In this study, we investigated the bacterial and fungal communities associated with different soil-plant compartments of two phylogenetically distant endemic plants, the Poaceae Poa kerguelensis and the Brassicaceae Pringlea antiscorbutica, in fellfields with contrasted pedoclimatic conditions. RESULTS: Using 16S rRNA gene and Internal Transcribed Spacer (ITS) region metabarcoding, we identified a strong soil-plant compartment effect affecting microbial communities, with bacterial and fungal α-diversity higher in bulk and rhizospheric soils and progressively decreasing in roots and above-ground compartments. The microbiota of the different soil-plant compartments studied differ in their recruitment patterns. The bacterial communities of the aerial parts of P. antiscorbutica were less dependent on those of the underground parts compared to those of P. kerguelensis. We also showed that the microbiota of distinct plant species and their different soil-plant compartments respond differently to pedoclimatic variables, with a greater impact of climatic variables over soil ones on aboveground bacterial microbiomes than on belowground microbiomes. CONCLUSIONS: Our results highlight the dual role of environmental variability and of the identity of the host on the recruitment and diversity of plant microbiomes in the isolated studied ecosystems. As plant holobionts are part of the global biogeochemical ecosystem functioning, our results suggest that plant species-specific microbial recruitment strategies and differential vulnerability to environmental factors should be included in predicting sub-Antarctic ecosystem response to global warming.