Determinism and stochasticity drive microbial community assembly and microbial interactions in calcareous glacier forefields

Abstract

Calcareous glacier forefields challenge prevailing ecological frameworks on microbial biodiversity and community assembly due to their unique bedrock. Early stages of soil development in these environments are notorious for their high turnover rates, demanding a high degree of replication for obtaining conclusive data. However, studies across different calcareous glaciers are still missing. Here, we robustly investigated both bacterial and fungal diversity, association networks, and assembly processes in four calcareous glacier forefields of the Alps, focusing on the earliest soil developmental stages (<25 years) early in the snow-free season. We found a diverse community of bacteria and fungi, potentially involved in P and N nutrient cycling. A core microbiome existing across all four locations suggests that certain microbes might be more successful colonizers of these ecosystems than others. Nearest taxon index revealed phylogenetically clustered microbial communities. These findings suggest that the distribution and colonization of some microbes were influenced by selective forces such as geography and climate during the early stages of soil development in calcareous glaciers. Interestingly, there were no common bacterial-fungal associations across the four locations, indicating that this habitat does not select for specific bacterial-fungal associations and that associations were driven by neutral processes. We discuss microbial communities and their interactions in these special calcareous glacier forefield habitats. Moreover, we present innovative approaches for studying microbial assembly that address both deterministic, intrinsic drivers, like specific microbial traits, and stochastic, extrinsic drivers, such as the opportunistic behavior of microbes.IMPORTANCEOur study is based on three fundamental and unique approaches: (i) we utilize the early stages of soil development in four glacier forefields across the Alpine range. This design implies high replicability in a natural setting, which is crucial for drawing general conclusions. (ii) Our study investigates glacier forefields with calcareous bedrock directly after snowmelt. These habitats and periods remain surprisingly underexplored. (iii) Our results underline the relevance of bacterial-fungal associations in microbial community assembly alongside dispersal, drift, and natural selection. Taken together, our study provides new insights into the development of complex microbial communities, their stabilization and predictability, including ecological implications.

Keywords: Alpine environment; abundance-occurrence relation; calcareous bedrock; enzymatic activity; microbial community assembly; neutral theory.

Fig 1

PCA of the four glacier forefields with respect to the environmental variables analyzed. Triangles indicate the soil properties for the plots of each location. Arrows indicate the contribution of each variable. Abbreviations: NO₃^-, nitrate; NH4+, ammonium; P, phosphorous; PO₄^3-, phosphate; TP, total phosphorus. Statistics: R² = 0.6787, F_3,36 = 25.35, P = 0.001.

Fig 2

Enzyme activities measured in the four glacier forefield soils. Values were normalized across all enzymes between 0 and 1 for readability.

Fig 3

α-Diversity across glacier forefields. (A and B) Shannon index of microbial communities detected at the different locations. Horizontal lines represent the median, while the boxes represent the inter-quartile range of the first and third quartiles. The vertical lines (whiskers) represent the maximal and minimal values. Points within each boxplot represent the means. Letters indicate differences between glaciers in bacterial (F_3,188 = 19.77, P < 0.001) and fungal (F_3,188 = 8.32, P < 0.001) diversity. (C and D) Venn diagrams including the number of bacterial OTUs that occurred in at least 10 samples (out of 50) in each location and fungal OTUs that occurred in at least five samples (out of 50) in each location. The number and percentage in the middle are the number of OTUs shared across all glaciers (core). (E and F) Relative abundances of bacterial and fungal genera belonging to the core of the Venn diagrams; relative abundances of unique bacterial and fungal genera are not shown; bacterial and fungal genera with relative abundances below 1% and 0.5%, respectively, were summarized in Others.

Fig 4

Community-level (β) diversity of bacterial and fungal communities among the glacier forefields. (A and B) NMDS of bacterial and fungal communities among locations (bacterial NMDS stress value = 0.114; PERMANOVA R² = 0.35, P = 0.001; fungal NMDS stress value = 0.158%; PERMANOVA R² = 0.19, P = 0.001). (C and D) NTI contributions of the bacterial and fungal communities for each glacier. Points represent NTI values for microbial communities within each sample plot, colored based on their difference to the expected value of zero (black, community significantly different, P < 0.05; gray, community not significantly different, P > 0.05; two-tailed t-test). Red points within each violin represent the means (n = NTI value scores), while the vertical lines represent the standard deviation. The kernel probability density of the data at different values is represented by the width of the violin graphs, that is, a wider part of the violin represents a higher density of scores at that particular level (i.e., for that location), while a narrower part of the violin represents a lower density of scores. Community (gray points) whose NTI is close to zero (horizontal dashed line) exhibits little phylogenetic clustering and indicates that the community is phylogenetically random.

Fig 5

(A) Estimation of the neutral processes for bacterial and fungal communities of the four glacier forefields based on the Sloan NCM. The species (OTUs = points) that occur more and less frequently than predicted are shown in orange and light blue, respectively. Dashed lines represent 95% confidence intervals, and the species (OTUs = points) falling within the confidence intervals (gray) are considered neutrally distributed. (B) NCM of each OTU (node) in pairwise associations of the calculated networks. Within each glacier forefield, the y-axes are composed of fungal and bacterial nodes (number of nodes on top of each axis) and contain the percentages of nodes that were assigned to a specific partition (colored strata) by the NCM. Flow segments connect a fungal node to its bacterial node as pairwise association predicted by the network.