Dispersal is one of the main processes shaping ecological communities. Yet, for species-rich communities in natural systems, the role of dispersal in community assembly remains relatively less studied compared to other processes. This is the case for fungal communities, for which predictable knowledge about where and how the dispersal propagules move across space is largely lacking. We sampled fungal communities at their dispersal stage in a lake mainland-island system in Finland, using a regular grid of 18 × 18 km, including sites on the mainland, islands and over the water. Fungal communities were screened by applying DNA barcoding to air samples. To assess the factors determining fungal dispersal, we modelled aerial fungal communities with a joint species distribution model, including spore traits, weather-related predictors, and spatial predictors. We found that the probability of occurrence of most species (and consequently species richness measured as the number of OTUs per sample) was lower in low-connectivity sites (water and isolated islands) compared to high-connectivity sites (mainland). There was a strong phylogenetic signal in how the fungal species responded to connectivity, indicating that some taxonomic groups are more dispersal limited than others, although such responses were not structured by their trophic guilds. Furthermore, wind speed influenced how species with different spore sizes responded to connectivity: in low-connectivity sites, species with large sexual spores were detected especially when wind was high, whereas, in high-connectivity sites, they were detected especially when wind was low. This study demonstrates that air fungal dispersal might be more predictable than previously considered and contributes to the mechanistic understanding of fungal air dispersal.
Read full abstract