A high degree of endemism in fungal communities suggests that fungal species may be adapted to relatively narrow envelopes of environmental conditions and plant communities. However, it remains unclear how these environmental and biotic factors structure fungal communities on a continental scale, especially across multiple years. This is important for forecasting the impacts of climate change on biodiversity and microbially-mediated ecosystem services. The NSF National Ecological Observatory Network (NEON) provides a public dataset containing ITS marker-gene amplicon sequences from across 47 terrestrial sites spanning all 20 ecoclimatic domains in the United States; from each site, up to 10 plots are randomly sampled at a frequency of 1-4 three times per year since 2014. We combine the NEON sampling network with a previous NSF Dimensions of Biodiversity project on North American soil fungi (DoB-FUN), which consisted of soil samples from ~70 mature pine forest sites spanning the continent. We leverage the combined dataset to compare spatial and temporal turnover rates in fungal communities, and to disentangle the effects of regional climate and local soil chemistry on fungal communities. In particular, we use generalized dissimilarity modeling (GDM) to account for non-linear responses to environmental gradients.
Results/Conclusions
Initial processing of the ITS marker-gene sequences yielded nearly 200,000 amplicon sequence variants (ASVs) across over 2,100 soil samples collected over 5 years. Across the dataset, fungal communities display a high degree of endemism, with a significant association between Bray-Curtis dissimilarity and geographic distance (GDM, P < 0.01) that saturates at approximately 1,000 km. Mean annual temperature is a strong driver of fungal community composition (P = 0.02), whereas differences in mean annual precipitation are important only among sites with low precipitation (P = 0.08). At local scales, soil pH is a strong driver of fungal community composition (P < 0.01). Fungal community composition varies substantially over time, but inter-annual and seasonal patterns are difficult to discern. These results support the hypothesis that climate is a strong driver of fungal community composition, and that projected changes in climate may have consequences for fungal species distributions. However, fine-scale community assembly will depend on soil pH. These findings demonstrate the utility of large-scale sampling networks in resolving nonlinear fungal community responses to climate change, and lay the groundwork for inferring how fungal species in regional pools jointly influence each other’s occurrences at local scales.