Do fungal communities mediate tree species interactions in mixed beech–conifer forests?

Understanding how tree species interact within forest stands and how these interactions influence forest functioning—particularly biotic interactions with soil microbes—is crucial for informing forest management strategies under a rapidly changing climate.

Altering tree species composition can shift soil fungal community structure and affect tree performance and competitive outcomes. For instance, enriching monospecific beech forests with conifers (mainly spruce or the non-native Douglas fir) can improve beech growth and stress resistance, especially under drought conditions, and it has more recently been linked to higher fungal diversity. Although several mechanisms have been proposed to explain these positive effects on beech growth, site conditions likely play a major role, and potentially also feedback with fungi, such as mycorrhizal symbionts.

Here, we investigated whether belowground fungal communities may act as mediators of tree species interactions in mixed beech–conifer forests. Specifically, we hypothesized that variation in fungal community composition is associated with variation in the intensity of tree species interactions, focusing on European beech. To show how beech growth differs under interspecific competition in beech-spruce and beech-Douglas fir forests, we calculated the relative interaction index (RII). We further hypothesize that the effects of site conditions on beech RII are indirect and mediated by fungal communities.

We calculated diameter increment of beech trees between 2017 and 2024 in pure beech stands and in mixed beech–spruce and beech–Douglas fir stands. Tree growth was estimated using allometric equations to derive annual aboveground biomass increment, which was used as the performance metric for calculating beech RII in mixed stands with either spruce or Douglas fir as competitors.

Soil- and root-associated fungal communities were characterized using DNA metabarcoding. Fungal community composition was analysed separately for soil and root samples using principal coordinates analysis (PCoA), and it was used to predict RII while accounting for the effects of site and stand covariates (e.g., stand age, stand density, and soil properties). To disentangle the relationships among soil environment, fungal community composition, and beech RII, we applied a stepwise regression framework reflecting a hypothesized causal pathway. We further examined associations between beech RII and differentially abundant fungal taxa putatively involved in mediating tree species interactions.

We found that beech RII was associated with fungal community composition but only in beech–spruce forests, indicating a strong neighbour identity effect. In beech–spruce stands, the influence of site conditions on beech RII was mediated by both soil and root fungal communities. Additionally, ectomycorrhizal fungal taxa which significantly differed in relative abundance between beech–spruce and pure beech forests were negatively correlated with beech RII, making them candidates involved in or responding to shifts in tree species interactions.

Overall, our results demonstrate that fungal communities are tightly coupled to tree species interactions in beech–spruce forests but not in beech–Douglas fir forests, where alternative mechanisms beyond soil conditions may predominantly regulate tree interactions.