Extreme summers impact cropland and grassland soil microbiomes

Abstract

The increasing frequency of extreme weather events highlights the need to understand how soil microbiomes respond to such disturbances. Here, metagenomics was used to investigate the effects of future climate scenarios (+0.6 °C warming and altered precipitation) on soil microbiomes during the summers of 2014-2019. Unexpectedly, Central Europe experienced extreme heatwaves and droughts during 2018-2019, causing significant impacts on the structure, assembly, and function of soil microbiomes. Specifically, the relative abundance of Actinobacteria (bacteria), Eurotiales (fungi), and Vilmaviridae (viruses) was significantly increased in both cropland and grassland. The contribution of homogeneous selection to bacterial community assembly increased significantly from 40.0% in normal summers to 51.9% in extreme summers. Moreover, genes associated with microbial antioxidant (Ni-SOD), cell wall biosynthesis (glmSMU, murABCDEF), heat shock proteins (GroES/GroEL, Hsp40), and sporulation (spoIID, spoVK) were identified as potential contributors to drought-enriched taxa, and their expressions were confirmed by metatranscriptomics in 2022. The impact of extreme summers was further evident in the taxonomic profiles of 721 recovered metagenome-assembled genomes (MAGs). Annotation of contigs and MAGs suggested that Actinobacteria may have a competitive advantage in extreme summers due to the biosynthesis of geosmin and 2-methylisoborneol. Future climate scenarios caused a similar pattern of changes in microbial communities as extreme summers, but to a much lesser extent. Soil microbiomes in grassland showed greater resilience to climate change than those in cropland. Overall, this study provides a comprehensive framework for understanding the response of soil microbiomes to extreme summers.

Fig. 1. GCEF, weather, and soil conditions during the summers of 2014–2019.

A Aerial photo of the GCEF experimental site in Bad Lauchstädt, Germany. Experimental design: two climate treatments (A and F) and three land-use types (CF, OF, and IG). B Monthly ambient precipitation and air temperature during the experiment. C Physicochemical properties of surface soil samples. Error bars represent the standard error of the mean. Significant effects of extreme summers, climate, and land-use changes on the soil properties were determined using linear mixed-effects models, significance levels: *p < 0.05; **p < 0.01; ***p < 0.001. GCEF Global Change Experimental Facility, CF conventional farming, OF organic farming, IG intensive grassland, MOI soil moisture, TC total carbon, TN total nitrogen, SMN soil mineral nitrogen. Photo: Tricklabor/Service Drohne.

Fig. 2. Impact of climate change on the composition of soil microbiomes during the summers of 2014–2019.

A, B Bar plots depict the proportion of major phyla in prokaryotic and fungal communities based on metagenomic reads, respectively. C Bar plot showing the proportion of major viral families in soil samples. Taxonomic groups with a relative abundance <0.5% were combined into “Others”. Asterisks indicate significant differences between normal and extreme summers based on DESeq2 (*p < 0.05; **p < 0.01; ***p < 0.001). D Mantel’s correlation analysis between soil properties and soil microbial communities. The Mantel test was performed at genus level (archaea, bacteria, and fungi), vOTU (DNA viruses), and KEGG KO (function) level, respectively. Pearson’s correlations were calculated between soil properties.

Fig. 3. Phylogenetic placement of dereplicated MAGs from this study.

The maximum likelihood tree of MAGs was constructed using RAxML based on conserved marker genes in CheckM. Bootstrap values > 90% are shown. Taxonomic classification of the MAGs was inferred using GTDB-tk. The heatmap indicates the abundance of MAGs in metagenomics (RPKM) during the summers of 2014–2019. The bar chart shows the abundance of MAGs in metatranscriptomics (RPKM) in July 2022. The completeness of MAGs is represented by the pies at the branch tips.

Fig. 4. Impact of climate and land-use changes on the microbial community assembly and function in cropland and grassland.

A Changes of homogeneous selection to bacterial community assembly were estimated by iCAMP. Data are presented as mean values ± SD. Error bars represented standard deviations (n = 15, 20, and 10 biologically independent samples for A/F, cropland, and grassland treatment, respectively). One-side significance based on bootstrapping test is expressed as ***p < 0.01. B Volcano plots depicting the abundances of genes (KEGG KO) that exhibit significant differences between the compared groups. Significantly altered genes are marked in red (up) and blue (down) based on DESeq2 p < 0.05.

Fig. 5. Cluster analysis of gene abundance profiles during the summers of 2014–2019.

A Six clusters show the abundance patterns of KEGG KOs across 6 years and the number of KOs in each cluster is displayed. B Heatmap showing KEGG modules (completeness > 60%) based on KOs in Cluster 1 and 2 (membership > 0.5). The RPKM value of each module was log₁₀ transformed before row scaling. CMP-KDO cytidine 5′-monophospho-3-deoxy-d-manno-2-octulosonic acid, PPP pentose phosphate pathway, UDP-GlcNAc uridine diphosphate N-acetylglucosamine, IMP inosine monophosphate, ADP adenosine diphosphate, ATP adenosine triphosphate.

Fig. 6. Impact of climate change on genes involved in microbial biogeochemical cycles and stress response.

Heatmaps displaying abundances of genes related to C, S, and P cycles (A), N cycle (B), CAZymes (C), and stress response (D) that were affected by climate change. Asterisks indicate significant differences between A and F treatments over 6 years based on DESeq2 (*p < 0.05; **p < 0.01; ***p < 0.001). The RPKM value of genes was log₁₀ transformed before row scaling. The RNA/DNA ratio of selected genes was calculated using metagenomics and metatranscriptomics. CBB Calvin–Benson–Bassham, rTCA reductive tricarboxylic acid, 3-HP 3-hydroxypropionate, 3-HP/4-HB 3-hydroxypropionate/4-hydroxybutyrate, ASR assimilatory sulfate reduction, DSR dissimilatory sulfate reduction, SOX sulfur oxidizing, SOD superoxide dismutase, CAT catalase, Hsp heat shock protein, GPx glutathione peroxidase.

Fig. 7. Impact of climate change on microbial BGCs during the summers of 2014–2019.

A Heatmaps showing abundances (>0.5%) of BGCs in metagenomic contigs significantly affected by extreme summers. The RPKM value of BGCs was log₁₀ transformed before row scaling. Asterisks indicate significant differences between A and F treatments based on the DESeq2 (*p < 0.05; **p < 0.01; ***p < 0.001). B Chemical structure of geosmin (GSM), 2-methylisoborneo (MIB), and alkylresorcinols. R represents alkyl side chains. C Phylum-level assignment of BGC-containing contigs. Numbers in brackets represent the number of contigs assigned to the same BGC. D BGCs in assembled MAGs as predicted by antiSMASH. The sequence similarities between query and reference BGCs in the MIBiG database are shown in brackets. PKS polyketide synthase, NRPS nonribosomal peptide synthetase.