Microbial growth under drought is confined to distinct taxa and modified by potential future climate conditions

Abstract

Climate change increases the frequency and intensity of drought events, affecting soil functions including carbon sequestration and nutrient cycling, which are driven by growing microorganisms. Yet we know little about microbial responses to drought due to methodological limitations. Here, we estimate microbial growth rates in montane grassland soils exposed to ambient conditions, drought, and potential future climate conditions (i.e., soils exposed to 6 years of elevated temperatures and elevated CO₂ levels). For this purpose, we combined ¹⁸O-water vapor equilibration with quantitative stable isotope probing (termed 'vapor-qSIP') to measure taxon-specific microbial growth in dry soils. In our experiments, drought caused >90% of bacterial and archaeal taxa to stop dividing and reduced the growth rates of persisting ones. Under drought, growing taxa accounted for only 4% of the total community as compared to 35% in the controls. Drought-tolerant communities were dominated by specialized members of the Actinobacteriota, particularly the genus Streptomyces. Six years of pre-exposure to future climate conditions (3 °C warming and + 300 ppm atmospheric CO₂) alleviated drought effects on microbial growth, through more drought-tolerant taxa across major phyla, accounting for 9% of the total community. Our results provide insights into the response of active microbes to drought today and in a future climate, and highlight the importance of studying drought in combination with future climate conditions to capture interactive effects and improve predictions of future soil-climate feedbacks.

Fig. 1. Effect of drought and future climate conditions on the composition of total and growing bacterial and archaeal communities.

Principal component analysis (PCA) of total (a) and growing (b) bacterial and archaeal communities (n = 4 replicates) on centered log-ratio transformed amplicon sequence variant (ASV) absolute abundances. Absolute abundances were calculated by multiplying ASV-specific amplicon sequencing reads with 16 S rRNA gene copies inferred by digital droplet PCR. Absolute abundances were agglomerated over the density fractions of a sample gradient. Statistics from two-way permutation-based multivariate analysis of variance (PERMANOVA) testing a full two factorial design (Drought _Yes, Drought _No, Climate _Ambient, Climate _Future) are provided as inset panels. Source data are provided with this paper.

Fig. 2. Diversity, population size, and mean relative growth rates of growing communities of bacteria and archaea across treatments.

a Number of unique growing amplicon sequence variants (ASVs) (n = 4 replicates). b Population size of the growing community expressed as the percentage of growing taxa of the total community based on the sum of their 16 S rRNA gene copies (n = 4 replicates). c Mean taxon-level relative growth rates (n = 4 means of taxon-specific growth rates calculated per replicate). Asterisks depict significant results from either two-way ANOVA testing a full two factorial design (a, b: Drought _Yes, Drought _No, Climate _Ambient, Climate _Future) or a linear mixed model (c, n = 391–1845 taxon-specific growth rates per replicate). Light-colored dots represent replicate samples (a, b) or taxon-specific growth rates (c) across treatment replicates, used to calculate mean taxon-level relative growth rates and standard errors (points, error bars). Boxes show the interquartile range with a line representing the median and minimum/maximum whiskers (a). Error bars represent standard errors (b, c). Source data are provided with this paper.

Fig. 3. Overlap of growing ASVs between treatments and growth response of shared taxa.

a Venn diagram of shared and unique amplicon sequence variants (ASVs) that were actively growing in at least two replicates per treatment. b Growth rates of shared taxa across four treatment comparisons: Ambient and Ambient + Drought (left panel), Future Climate and Future Climate + Drought (left center panel), Ambient + Drought and Future Climate + Drought (right center panel), Ambient and Future Climate (right panel). Asterisks represent significant differences between comparisons using two-sided Student’s t-tests (n = 4 means of taxon-specific growth rates calculated per replicate). Light-colored dots represent taxon-specific growth rates across replicates used to calculate means and standard errors (points, error bars). Source data are provided with this paper.

Fig. 4. Relative changes in the number of growing taxa per phylum under drought and future drought conditions.

a Relative change (0 = no change, 1 = increase by 100%) in the number of growing taxa per phylum comparing Ambient (reference) and Ambient + Drought conditions (n = 4 replicates). b Relative change in the number of growing taxa comparing Ambient + Drought (reference) and Future Climate + Drought conditions (n = 4 replicates). Points represent mean relative changes in the number of growing taxa per phylum including standard errors (error bars). Two-way ANOVA testing a full two factorial design (Drought _Yes, Drought _No, Climate _Ambient, Climate _Future) was performed on the number of growing amplicon sequence variants (ASVs) across all treatments for each phylum independently that fulfilled normality and homoscedasticity requirements. P-values were adjusted for multiple testing using false discovery rate correction. Bold colored points represent significant effects (a: Drought = p < 0.05, b: Climate = p < 0.05 or Drought x Climate = p < 0.05) and transparent points represent non-significant effects. Vertical lines depict the means of the respective reference treatment (n = 4 replicates) used to calculate relative changes. Source data and p-values are provided with this paper.

Fig. 5. Drought-induced changes in the top ¹⁸O assimilating taxa agglomerated at the genus level.

Heatmap showing taxa with the highest proportional ¹⁸O assimilation (contribution to the total community’s growth) under ambient precipitation (a) and drought (b), visualized across all treatments and individual samples (rectangles, n = 4). Amplicon sequence variants (ASVs) were ranked based on their proportional ¹⁸O assimilation, separately, for drought-unaffected (a: Ambient, Future Climate) and drought-affected samples (b: Ambient + Drought, Future Climate + Drought). The top five ASVs per sample were then selected (a: 26 total unique ASVs; b: 19 total unique ASVs) and visualized. Proportional ¹⁸O assimilation ranges from 0−1 and estimates how much a single taxon contributes to the community’s overall growth. It is calculated using re-computed relative abundances of only growing taxa (sum of growing taxa = 1) and their relative growth rates (RGR). ASVs had to be active in at least two samples if detected as growing in a treatment. ASV identities were agglomerated at the genus level and sorted in descending order based on proportional ¹⁸O assimilation. If genus identity could not be assigned (NA), we agglomerated taxa at the family or phylum level. Source data are provided with this paper.