Opportunistic partner choice among arctic plants and root-associated fungi is driven by environmental conditions.
Interactions between plants and soil microbes play an important role in structuring plant communities. Yet, little is known about how fungal networks are structured on the one hand by fungal responses to their environment (including their host plant) and on the other by responses to each other. We quantified changes in plant-fungus networks along geographic and environmental gradients across the Arctic, assessing the degree to which plants and fungi showed preference for specific interaction partners and how specificity varies along environmental gradients. To this aim, we sampled roots of 12 widely distributed plant taxa: Saxifraga oppositifolia; Bistorta vivipara; Dryas spp.; Vaccinium tis-idaea; Vaccinium uliginosum; Vaccinium myrtillus; Empetrum nigrum; Betula nana; alix arctica; Salix polaris; Cassiope tetragona; and Silene acaulis. To quantify the pool of fungi from which plant roots may recruit association partners, we also sampled fungi in the surrounding soil. Identifying fungaI communities by DNA metabarcoding, we used Hierarchical Modelling of Species Communities (HMSC) to assess how fungal communities change along environmental gradients, and whether plants actively select their root-associated fungi from the pool of fungi present in the bulk soil. We found that although the fungal communities within the soil and rhizosphere share 85% of genera, their composition differs significantly from each other. The two community types show similar responses to the environment and taxa show low partner fidelity. Thus, the structure of fungal communities on plant rhizosphere is mainly driven by abiotic rather than biotic conditions. Overall, in comparison with null models, networks of plants and rhizosphere-associated fungi showed a distinctly non-random structure, responding strongly to pH and temperature gradients. Our findings suggest that the dynamics and structure of plant-root associated interactions might be severely altered by abiotic changes in the rapidly changing arctic environment. Open Research statementData are privately provided for peer review. The raw sequences for the soil and root samples generated during the current study will be available in the Sequence Read Archive repository, in the BioProject PRJNA1094865 upon acceptance. For review purposes, the code and datasets used for the analyses of this study are temporarily available in Figshare open access repository at https://figshare.com/s/1b074f1751682d3487cf upon acceptance.