Mechanisms of fungal community assembly in wild stoneflies moderated by host characteristics and local environment

Abstract

Deterministic and stochastic forces both drive microbiota assembly in animals, yet their relative contribution remains elusive, especially in wild aquatic-insect-associated fungal communities. Here, we applied amplicon sequencing to survey the assembly mechanisms of the fungal community in 155 wild stonefly individuals involving 44 species of 20 genera within eight families collected from multiple locations in China. Analysis showed that fungal diversity and network complexity differed significantly among the eight stonefly families, and that the fungal communities in stoneflies exhibited a significant distance-decay pattern across large spatial scales. Both a structural equation model and variance partitioning analysis revealed that environmental factors (e.g., geographical, climatic) outweigh host attributes in shaping the fungal community of stoneflies. Using neutral and null model analyses, we also find that deterministic processes play a larger role than stochasticity in driving the fungal community assembly. However, the relative contribution of ecological processes including dispersal, drift, and selection, varied strongly with host taxonomy. Furthermore, environmental conditions also significantly affect the strength of these ecological processes. Overall, our findings illustrate that variations in host attributes and environment factors may moderate the relative influence of deterministic and stochastic processes to fungal community composition in wild stoneflies, which provides new insights into mechanisms of microbial community assembly in aquatic arthropods.

Fig. 1. Sampling locations of wild stoneflies across multiple geographic regions of China.

Differently colored symbols represent different stonefly families. The figure was generated using ArcGIS 10 Crack software based on a template map from the Chinese National Basic Geographic Information Center (http://ngcc.sbsm.gov.cn).

Fig. 2. Fungal community diversity and composition among eight stonefly families.

a Richness and Shannon indices of different stonefly fungal communities. Different letters denote significant differences between stonefly families with ANOVA tests (p < 0.05). b Principal coordinate analysis (PCoA) of fungal community Bray-Curtis dissimilarities with permutational analysis of variance among different stonefly families. The variation explained by each principal coordinate is denoted in parentheses. c Relative abundances of each fungal genus in different stonefly families. d Co-occurrence networks of the fungal communities from each stonefly family based on Spearman’s correlation analysis between OTUs. Blue and red lines represent significant negative and positive correlations, respectively. The sizes of the points indicate the relative abundance of OTUs in each microbial community.

Fig. 3. Geographic patterns in stonefly fungal communities.

Distance-decay curves of fungal communities based on Bray–Curtis similarities in wild populations of stoneflies. The line represents the ordinary least-squares linear regression.

Fig. 4. The effect of host-related and environmental factors on the stonefly fungal community composition.

a Structural equation model provided insight on connections between both host-related and environmental variables to fungal community composition. The blue and red arrows indicate statistically significant negative and positive paths, respectively. The width of the arrows represents the strengths of these relationships. The R² values under each box indicate the amount of variation in that variable explained by the input arrows. Numbers next to arrows are unstandardized slopes. b Variance partitioning analysis (VPA).

Fig. 5. Fit of a neutral model on fungal community assembly in stoneflies.

The yellow solid and dashed lines indicate the predicted occurrence and 95% confidence interval of the neutral model, respectively. R² indicates the goodness of fit to the neutral model, and m shows the migration rate.

Fig. 6. Mechanisms of stonefly fungal community assembly evaluated using null model analysis.

Contributions of deterministic (|βNTI | ≥ 2) and stochastic processes (|βNTI | < 2) on fungal community assembly in all stoneflies (a) and in each of eight stonefly families (c). The relative contribution of ecological processes (i.e., homogeneous selection, heterogeneous selection, homogenizing dispersal, dispersal limitation, and drift) in driving the fungal assembly in all samples (b), and in each stonefly family (d).

Fig. 7. Effects of environmental factors on stonefly fungal community assembly.

Mantel analysis used to evaluate the correlation between the β-Nearest Taxon Index (βNTI) and the multiples environmental variables: (a) altitude, (b) latitude, (c) longitude, (d) annual mean temperature (AMT) and (e) annual mean precipitation (AP).