High throughput sequencing combined with null model tests reveals specific plant-fungi associations linked to seedling establishment and survival

1. Plant-fungal interactions are important for plant community assembly, but quantifying these relationships remains challenging. High throughput sequencing of fungal communities allows us to identify plant-fungal associations at a high level of resolution, but often fails to provide information on taxonomic and functional assignment of fungi. 2. We transplanted seeds of Pinus cembra a cross an elevational gradient (1850-2250 m a.s.l.) and identified environmental factors and known fungal associates important for seedling establishment and survival. We then applied null model tests to identify taxonomically unassigned fungi associated with pine recruitment. 3. Early seedling establishment was determined by abiotic environmental factors, while seedling survival was predominantly affected by biotic environmental factors (i.e., the abundance of a fungal pathogen known from literature and the distance to adult trees). Null model tests identified known mycorrhizal partners and a large number of unknown operational taxonomic units (OTUs) associated with seedling survival, including saprotrophic and pathogenic species. These results highlight that unknown fungal OTUs, which are usually discarded from analyses, could play a crucial role for plant survival. 4. Synthesis. We conclude that high throughput metabarcoding paired with null model tests, is a valuable approach for identifying hidden plant-fungal associations within large and complex DNA metabarcoding datasets. Such an approach can be an important tool in illuminating the black box of plant-microbe interactions, and thus understanding ecosystem dynamics.